From 535759239ceec08230027dd1cb377dcc215cb7fb Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Pietil=C3=A4?= <> Date: Tue, 3 Feb 2026 10:16:29 -0500 Subject: [PATCH] Use instance strings in configs. --- .../bwd_data/profiler/ndhwgc_bf16.conf | 164 +++--- .../bwd_data/profiler/ndhwgc_fp16.conf | 164 +++--- .../bwd_data/profiler/ndhwgc_fp32.conf | 140 ++--- .../configs/bwd_data/profiler/nhwgc_bf16.conf | 164 +++--- .../configs/bwd_data/profiler/nhwgc_fp16.conf | 164 +++--- .../configs/bwd_data/profiler/nhwgc_fp32.conf | 140 ++--- .../bwd_weight/profiler/ndhwgc_bf16.conf | 466 ++++++++--------- .../bwd_weight/profiler/ndhwgc_fp16.conf | 466 ++++++++--------- .../bwd_weight/profiler/ndhwgc_fp32.conf | 88 ++-- .../bwd_weight/profiler/nhwgc_bf16.conf | 480 ++++++++--------- .../bwd_weight/profiler/nhwgc_fp16.conf | 486 +++++++++--------- .../bwd_weight/profiler/nhwgc_fp32.conf | 88 ++-- 12 files changed, 1505 insertions(+), 1505 deletions(-) diff --git a/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/ndhwgc_bf16.conf b/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/ndhwgc_bf16.conf index 4ee0de66d1..e2ee34b27b 100644 --- a/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/ndhwgc_bf16.conf +++ b/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/ndhwgc_bf16.conf @@ -1,82 +1,82 @@ -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Default, 32, 32, 2, 2, 1, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Default, 32, 32, 1, 2, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Default, 32, 32, 2, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Default, 32, 32, 2, 4, 8, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 256, 128, 32, 8, 8, Default, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Default, 32, 32, 2, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 128, 32, 8, 8, Default, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 128, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 64, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 64, 128, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 64, 32, 8, 8, Default, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Default, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Default, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 32, 128, 32, 8, 8, Default, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 32, 32, 8, 8, Default, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 32, 64, 32, 8, 8, Default, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 256, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 32, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 32, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Default, 32, 32, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Default, 32, 32, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Default, 32, 32, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Default, 16, 16, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Default, 16, 16, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Default, 16, 16, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Filter1x1Stride1Pad0, 32, 32, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Filter1x1Stride1Pad0, 32, 32, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Filter1x1Stride1Pad0, 32, 32, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Filter1x1Stride1Pad0, 16, 16, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Filter1x1Stride1Pad0, 16, 16, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Filter1x1Stride1Pad0, 16, 16, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 1, 1, 1> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/ndhwgc_fp16.conf b/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/ndhwgc_fp16.conf index 4ee0de66d1..e2ee34b27b 100644 --- a/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/ndhwgc_fp16.conf +++ b/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/ndhwgc_fp16.conf @@ -1,82 +1,82 @@ -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Default, 32, 32, 2, 2, 1, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Default, 32, 32, 1, 2, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Default, 32, 32, 2, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Default, 32, 32, 2, 4, 8, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 256, 128, 32, 8, 8, Default, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Default, 32, 32, 2, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 128, 32, 8, 8, Default, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 128, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 64, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 64, 128, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 64, 32, 8, 8, Default, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Default, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Default, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 32, 128, 32, 8, 8, Default, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 32, 32, 8, 8, Default, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 32, 64, 32, 8, 8, Default, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 256, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 32, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 32, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Default, 32, 32, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Default, 32, 32, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Default, 32, 32, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Default, 16, 16, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Default, 16, 16, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Default, 16, 16, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Filter1x1Stride1Pad0, 32, 32, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Filter1x1Stride1Pad0, 32, 32, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Filter1x1Stride1Pad0, 32, 32, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Filter1x1Stride1Pad0, 16, 16, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Filter1x1Stride1Pad0, 16, 16, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Filter1x1Stride1Pad0, 16, 16, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 1, 1, 1> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/ndhwgc_fp32.conf b/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/ndhwgc_fp32.conf index 0391d33eb7..da90269493 100644 --- a/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/ndhwgc_fp32.conf +++ b/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/ndhwgc_fp32.conf @@ -1,70 +1,70 @@ -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Default, 32, 32, 2, 2, 1, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Default, 32, 32, 1, 2, 1, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Default, 32, 32, 2, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Default, 32, 32, 2, 4, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 256, 128, 32, 8, 8, Default, 32, 32, 4, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Default, 32, 32, 2, 4, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 128, 32, 8, 8, Default, 32, 32, 4, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 128, 32, 8, 8, Default, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 64, 32, 8, 8, Default, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 64, 128, 32, 8, 8, Default, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Default, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 64, 32, 8, 8, Default, 32, 32, 2, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Default, 32, 32, 1, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Default, 32, 32, 2, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 32, 128, 32, 8, 8, Default, 32, 32, 1, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 32, 32, 8, 8, Default, 32, 32, 2, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 32, 64, 32, 8, 8, Default, 32, 32, 1, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 1, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 4, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 256, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 4, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 4, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 4, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 32, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 32, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 1, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 1, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 1, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 1, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 1, 1, 1> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/nhwgc_bf16.conf b/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/nhwgc_bf16.conf index 4ee0de66d1..e2ee34b27b 100644 --- a/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/nhwgc_bf16.conf +++ b/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/nhwgc_bf16.conf @@ -1,82 +1,82 @@ -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Default, 32, 32, 2, 2, 1, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Default, 32, 32, 1, 2, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Default, 32, 32, 2, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Default, 32, 32, 2, 4, 8, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 256, 128, 32, 8, 8, Default, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Default, 32, 32, 2, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 128, 32, 8, 8, Default, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 128, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 64, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 64, 128, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 64, 32, 8, 8, Default, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Default, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Default, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 32, 128, 32, 8, 8, Default, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 32, 32, 8, 8, Default, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 32, 64, 32, 8, 8, Default, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 256, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 32, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 32, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Default, 32, 32, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Default, 32, 32, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Default, 32, 32, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Default, 16, 16, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Default, 16, 16, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Default, 16, 16, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Filter1x1Stride1Pad0, 32, 32, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Filter1x1Stride1Pad0, 32, 32, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Filter1x1Stride1Pad0, 32, 32, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Filter1x1Stride1Pad0, 16, 16, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Filter1x1Stride1Pad0, 16, 16, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Filter1x1Stride1Pad0, 16, 16, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 1, 1, 1> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/nhwgc_fp16.conf b/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/nhwgc_fp16.conf index 4ee0de66d1..e2ee34b27b 100644 --- a/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/nhwgc_fp16.conf +++ b/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/nhwgc_fp16.conf @@ -1,82 +1,82 @@ -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Default, 32, 32, 2, 2, 1, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Default, 32, 32, 1, 2, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Default, 32, 32, 2, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Default, 32, 32, 2, 4, 8, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 256, 128, 32, 8, 8, Default, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Default, 32, 32, 2, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 128, 32, 8, 8, Default, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 128, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 64, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 64, 128, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Default, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 64, 32, 8, 8, Default, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Default, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Default, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 32, 128, 32, 8, 8, Default, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 32, 32, 8, 8, Default, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 32, 64, 32, 8, 8, Default, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 256, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 32, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 32, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 1, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Default, 32, 32, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Default, 32, 32, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Default, 32, 32, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Default, 16, 16, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Default, 16, 16, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Default, 16, 16, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Filter1x1Stride1Pad0, 32, 32, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Filter1x1Stride1Pad0, 32, 32, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 64, 16, 16, Filter1x1Stride1Pad0, 32, 32, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Filter1x1Stride1Pad0, 16, 16, 1, 1, 16, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Filter1x1Stride1Pad0, 16, 16, 1, 1, 16, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 64, 16, 16, Filter1x1Stride1Pad0, 16, 16, 1, 1, 16, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 1, 1, 1> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/nhwgc_fp32.conf b/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/nhwgc_fp32.conf index 0391d33eb7..da90269493 100644 --- a/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/nhwgc_fp32.conf +++ b/experimental/grouped_convolution_tile_instances/configs/bwd_data/profiler/nhwgc_fp32.conf @@ -1,70 +1,70 @@ -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Default, 32, 32, 2, 2, 1, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Default, 32, 32, 1, 2, 1, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Default, 32, 32, 2, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Default, 32, 32, 2, 4, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 256, 128, 32, 8, 8, Default, 32, 32, 4, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Default, 32, 32, 2, 4, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 128, 32, 8, 8, Default, 32, 32, 4, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 128, 32, 8, 8, Default, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 64, 32, 8, 8, Default, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 64, 128, 32, 8, 8, Default, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Default, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 64, 32, 8, 8, Default, 32, 32, 2, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Default, 32, 32, 1, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Default, 32, 32, 2, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 32, 128, 32, 8, 8, Default, 32, 32, 1, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 32, 32, 8, 8, Default, 32, 32, 2, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 32, 64, 32, 8, 8, Default, 32, 32, 1, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 1, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 4, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 256, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 4, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 256, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 4, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 4, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<128, 32, 128, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 32, 64, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Default, 16, 16, 1, 4, 1, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Default, 16, 16, 4, 1, 1, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 16, 64, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 4, 1, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 4, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<64, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 4, 1, 1, 4, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Default, 32, 32, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Default, 32, 32, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Default, 16, 16, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Default, 16, 16, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 32, 8, 8, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 128, 32, 16, 4, 4, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 8, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 2, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 32, 8, 8, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 1, 1, 1> -DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1<256, 64, 16, 16, 4, 4, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 1, 1, 1> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/ndhwgc_bf16.conf b/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/ndhwgc_bf16.conf index 39dca09511..691d6e595e 100644 --- a/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/ndhwgc_bf16.conf +++ b/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/ndhwgc_bf16.conf @@ -1,233 +1,233 @@ -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Default, 8, 2, 2, 2, 4, 2, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Default, 8, 2, 1, 8, 4, 2, 1, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Default, 8, 1, 2, 2, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 256, 128, 4, Default, 8, 4, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 256, 4, Default, 8, 2, 4, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 128, 4, Default, 8, 4, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 128, 4, Default, 8, 2, 2, 8, 2, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 64, 4, Default, 8, 2, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 64, 128, 4, Default, 8, 2, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Default, 8, 2, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 64, 4, Default, 8, 2, 1, 8, 2, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 128, 4, Default, 8, 1, 2, 8, 1, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Default, 8, 2, 1, 8, 4, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 32, 128, 4, Default, 8, 1, 2, 8, 1, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 32, 4, Default, 8, 2, 1, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Default, 8, 1, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Filter1x1Stride1Pad0, 8, 2, 2, 2, 4, 2, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 4, 2, 1, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Filter1x1Stride1Pad0, 8, 1, 2, 2, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 256, 128, 4, Filter1x1Stride1Pad0, 8, 4, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 256, 4, Filter1x1Stride1Pad0, 8, 2, 4, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 128, 4, Filter1x1Stride1Pad0, 8, 4, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 128, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 2, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 64, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 64, 128, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 64, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 2, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 128, 4, Filter1x1Stride1Pad0, 8, 1, 2, 8, 1, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 4, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 32, 128, 4, Filter1x1Stride1Pad0, 8, 1, 2, 8, 1, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 32, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Filter1x1Stride1Pad0, 8, 1, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Filter1x1Stride1Pad0, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Filter1x1Stride1Pad0, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Filter1x1Stride1Pad0, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Filter1x1Stride1Pad0, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Filter1x1Stride1Pad0, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Filter1x1Stride1Pad0, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Filter1x1Stride1Pad0, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Filter1x1Stride1Pad0, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Filter1x1Stride1Pad0, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Filter1x1Stride1Pad0, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Filter1x1Stride1Pad0, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Filter1x1Stride1Pad0, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Filter1x1Stride1Pad0, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Filter1x1Stride1Pad0, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 8, 8, 8, 8, 1, 1, 4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 4, 8, 1, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 1, 8, 4, 8, 1, 1, 4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 2, 8, 2, 8, 1, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 1, 8, 1, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 256, 32, Default, 8, 1, 16, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 128, 32, Default, 8, 1, 8, 4, 8, 4, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 48, 64, 32, Default, 8, 3, 4, 3, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 48, 32, Default, 8, 4, 3, 4, 4, 3, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 208, 32, Default, 8, 4, 13, 4, 4, 13, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 48, 64, 32, Default, 8, 3, 4, 3, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 48, 32, Default, 8, 4, 3, 4, 4, 3, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 208, 32, Default, 8, 4, 13, 4, 4, 13, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 1x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,1,1,Seq(1,16,1,4),2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,2,1,true,1,1,Seq(1,32,1,4),2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,2,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,256,128,4,8,32,32,4,2,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,8),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,256,4,8,32,32,2,4,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,8),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,128,4,8,32,32,4,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,128,4,8,32,32,2,2,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,64,4,8,32,32,2,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,64,128,4,8,32,32,2,2,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,64,4,8,32,32,2,1,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,128,4,8,32,32,1,2,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,32,128,4,8,32,32,1,2,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,4,8,32,32,2,1,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,1,1,Seq(1,16,1,4),2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,2,1,true,1,1,Seq(1,32,1,4),2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,2,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,256,128,4,8,32,32,4,2,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,8),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,256,4,8,32,32,2,4,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,8),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,128,4,8,32,32,4,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,128,4,8,32,32,2,2,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,64,4,8,32,32,2,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,64,128,4,8,32,32,2,2,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,64,4,8,32,32,2,1,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,128,4,8,32,32,1,2,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,32,128,4,8,32,32,1,2,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,4,8,32,32,2,1,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,4,1),Seq(2,0,1),Seq(2,0,1),1,8,8,false,Seq(8,4,1),Seq(2,0,1),Seq(2,0,1),1,8,8,false,1,1,Seq(1,16,1,16),4,Intrawave,v1,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,16,1),Seq(2,0,1),Seq(2,0,1),1,4,8,false,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,1,1,Seq(1,4,1,64),1,Intrawave,v1,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,Seq(8,16,1),Seq(2,0,1),Seq(2,0,1),1,4,8,false,1,1,Seq(1,16,1,16),4,Intrawave,v1,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,2,8,false,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,2,8,false,1,1,Seq(1,8,1,32),2,Intrawave,v1,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,1,1,Seq(1,4,1,64),1,Intrawave,v1,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,256,32,8,16,16,1,16,Seq(4,2,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,4,1,16),1,Intrawave,v1,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,128,32,8,16,16,1,8,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,4,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,8,false,1,1,Seq(1,4,1,16),1,Intrawave,v1,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,48,64,32,8,16,16,3,4,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,48,32,8,16,16,4,3,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,208,32,8,16,16,4,13,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,13,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,48,64,32,8,16,16,3,4,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,48,32,8,16,16,4,3,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,208,32,8,16,16,4,13,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,13,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,bf16,bf16,1,1> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/ndhwgc_fp16.conf b/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/ndhwgc_fp16.conf index b27f05f10b..7d992ee232 100644 --- a/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/ndhwgc_fp16.conf +++ b/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/ndhwgc_fp16.conf @@ -1,233 +1,233 @@ -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Default, 8, 2, 2, 2, 4, 2, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Default, 8, 2, 1, 8, 4, 2, 1, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Default, 8, 1, 2, 2, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 256, 128, 4, Default, 8, 4, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 256, 4, Default, 8, 2, 4, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 128, 4, Default, 8, 4, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 128, 4, Default, 8, 2, 2, 8, 2, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 64, 4, Default, 8, 2, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 64, 128, 4, Default, 8, 2, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Default, 8, 2, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 64, 4, Default, 8, 2, 1, 8, 2, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 128, 4, Default, 8, 1, 2, 8, 1, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Default, 8, 2, 1, 8, 4, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 32, 128, 4, Default, 8, 1, 2, 8, 1, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 32, 4, Default, 8, 2, 1, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Default, 8, 1, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Filter1x1Stride1Pad0, 8, 2, 2, 2, 4, 2, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 4, 2, 1, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Filter1x1Stride1Pad0, 8, 1, 2, 2, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 256, 128, 4, Filter1x1Stride1Pad0, 8, 4, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 256, 4, Filter1x1Stride1Pad0, 8, 2, 4, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 128, 4, Filter1x1Stride1Pad0, 8, 4, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 128, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 2, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 64, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 64, 128, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 64, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 2, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 128, 4, Filter1x1Stride1Pad0, 8, 1, 2, 8, 1, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 4, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 32, 128, 4, Filter1x1Stride1Pad0, 8, 1, 2, 8, 1, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 32, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Filter1x1Stride1Pad0, 8, 1, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Filter1x1Stride1Pad0, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Filter1x1Stride1Pad0, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Filter1x1Stride1Pad0, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Filter1x1Stride1Pad0, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Filter1x1Stride1Pad0, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Filter1x1Stride1Pad0, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Filter1x1Stride1Pad0, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Filter1x1Stride1Pad0, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Filter1x1Stride1Pad0, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Filter1x1Stride1Pad0, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Filter1x1Stride1Pad0, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Filter1x1Stride1Pad0, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Filter1x1Stride1Pad0, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Filter1x1Stride1Pad0, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 8, 8, 8, 8, 1, 1, 4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 4, 8, 1, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 1, 8, 4, 8, 1, 1, 4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 2, 8, 2, 8, 1, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 1, 8, 1, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 256, 32, Default, 8, 1, 16, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 128, 32, Default, 8, 1, 8, 4, 8, 4, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 48, 64, 32, Default, 8, 3, 4, 3, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 48, 32, Default, 8, 4, 3, 4, 4, 3, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 208, 32, Default, 8, 4, 13, 4, 4, 13, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 48, 64, 32, Default, 8, 3, 4, 3, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 48, 32, Default, 8, 4, 3, 4, 4, 3, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 208, 32, Default, 8, 4, 13, 4, 4, 13, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 1x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,1,1,Seq(1,16,1,4),2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,2,1,true,1,1,Seq(1,32,1,4),2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,2,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,256,128,4,8,32,32,4,2,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,8),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,256,4,8,32,32,2,4,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,8),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,128,4,8,32,32,4,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,128,4,8,32,32,2,2,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,64,4,8,32,32,2,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,64,128,4,8,32,32,2,2,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,64,4,8,32,32,2,1,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,128,4,8,32,32,1,2,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,32,128,4,8,32,32,1,2,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,4,8,32,32,2,1,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,1,1,Seq(1,16,1,4),2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,2,1,true,1,1,Seq(1,32,1,4),2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,2,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,256,128,4,8,32,32,4,2,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,8),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,256,4,8,32,32,2,4,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,8),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,128,4,8,32,32,4,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,128,4,8,32,32,2,2,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,64,4,8,32,32,2,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,64,128,4,8,32,32,2,2,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,64,4,8,32,32,2,1,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,128,4,8,32,32,1,2,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,32,128,4,8,32,32,1,2,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,4,8,32,32,2,1,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,8,1),Seq(2,0,1),Seq(2,0,1),1,8,8,false,Seq(8,8,1),Seq(2,0,1),Seq(2,0,1),1,8,8,false,1,1,Seq(1,16,1,16),4,Intrawave,v1,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,16,1),Seq(2,0,1),Seq(2,0,1),1,4,8,false,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,1,1,Seq(1,4,1,64),1,Intrawave,v1,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,Seq(8,16,1),Seq(2,0,1),Seq(2,0,1),1,4,8,false,1,1,Seq(1,16,1,16),4,Intrawave,v1,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,2,8,false,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,2,8,false,1,1,Seq(1,8,1,32),2,Intrawave,v1,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,1,1,Seq(1,4,1,64),1,Intrawave,v1,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,256,32,8,16,16,1,16,Seq(4,2,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,4,1,16),1,Intrawave,v1,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,128,32,8,16,16,1,8,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,4,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,8,false,1,1,Seq(1,4,1,16),1,Intrawave,v1,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,48,64,32,8,16,16,3,4,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,48,32,8,16,16,4,3,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,208,32,8,16,16,4,13,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,13,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,48,64,32,8,16,16,3,4,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,48,32,8,16,16,4,3,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,208,32,8,16,16,4,13,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,13,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,fp16,fp16,1,1> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/ndhwgc_fp32.conf b/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/ndhwgc_fp32.conf index b6386d41ad..4f3da2883a 100644 --- a/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/ndhwgc_fp32.conf +++ b/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/ndhwgc_fp32.conf @@ -1,44 +1,44 @@ -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Default, 4, 2, 2, 1, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Default, 4, 2, 1, 4, 4, 1, 1, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Default, 4, 1, 2, 1, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 256, 128, 4, Default, 4, 4, 2, 4, 4, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 256, 4, Default, 4, 2, 4, 4, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 128, 4, Default, 4, 4, 2, 4, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 128, 4, Default, 4, 2, 2, 4, 2, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 64, 4, Default, 4, 2, 2, 4, 4, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 64, 128, 4, Default, 4, 2, 2, 4, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Default, 4, 2, 2, 4, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 64, 4, Default, 4, 2, 1, 4, 2, 4, 1, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 128, 4, Default, 4, 1, 2, 4, 1, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Default, 4, 2, 1, 4, 4, 4, 1, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 32, 128, 4, Default, 4, 1, 2, 4, 1, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 32, 4, Default, 4, 2, 1, 4, 4, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Default, 4, 1, 2, 4, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Default, 8, 1, 1, 4, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Default, 8, 1, 1, 4, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Default, 8, 1, 1, 1, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Filter1x1Stride1Pad0, 4, 2, 2, 1, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Filter1x1Stride1Pad0, 4, 2, 1, 4, 4, 1, 1, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Filter1x1Stride1Pad0, 4, 1, 2, 1, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 256, 128, 4, Filter1x1Stride1Pad0, 4, 4, 2, 4, 4, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 256, 4, Filter1x1Stride1Pad0, 4, 2, 4, 4, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 128, 4, Filter1x1Stride1Pad0, 4, 4, 2, 4, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 128, 4, Filter1x1Stride1Pad0, 4, 2, 2, 4, 2, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 64, 4, Filter1x1Stride1Pad0, 4, 2, 2, 4, 4, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 64, 128, 4, Filter1x1Stride1Pad0, 4, 2, 2, 4, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Filter1x1Stride1Pad0, 4, 2, 2, 4, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 64, 4, Filter1x1Stride1Pad0, 4, 2, 1, 4, 2, 4, 1, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 128, 4, Filter1x1Stride1Pad0, 4, 1, 2, 4, 1, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Filter1x1Stride1Pad0, 4, 2, 1, 4, 4, 4, 1, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 32, 128, 4, Filter1x1Stride1Pad0, 4, 1, 2, 4, 1, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 32, 4, Filter1x1Stride1Pad0, 4, 2, 1, 4, 4, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Filter1x1Stride1Pad0, 4, 1, 2, 4, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Filter1x1Stride1Pad0, 8, 1, 1, 4, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Filter1x1Stride1Pad0, 8, 1, 1, 4, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Filter1x1Stride1Pad0, 8, 1, 1, 1, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Filter1x1Stride1Pad0, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 16, 16, 32, Filter1x1Stride1Pad0, 8, 1, 1, 1, 4, 1, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 16, 16, 32, Filter1x1Stride1Pad0, 8, 1, 1, 1, 4, 1, 4, 1, 1, 2> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,64,4,4,32,32,2,2,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,1,4,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,1,4,true,1,1,Seq(1,16,1,4),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,32,4,4,32,32,2,1,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,1,1,true,1,1,Seq(1,32,1,4),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,4,4,32,32,1,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,1,2,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,256,128,4,4,32,32,4,2,Seq(1,4,64,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,8),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,256,4,4,32,32,2,4,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,64,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,8),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,128,4,4,32,32,4,2,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,128,4,4,32,32,2,2,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,64,4,4,32,32,2,2,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,128,64,128,4,4,32,32,2,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,64,4,4,32,32,2,2,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,64,4,4,32,32,2,1,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,128,4,4,32,32,1,2,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,32,4,4,32,32,2,1,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,128,32,128,4,4,32,32,1,2,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,4,4,32,32,2,1,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,4,4,32,32,1,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,8,8,32,32,1,1,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,1,1,Seq(1,16,1,16),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,8,8,32,32,1,1,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,1,1,Seq(1,4,1,64),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,8,8,32,32,1,1,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,1,1,Seq(1,16,1,16),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,8,8,32,32,1,1,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,1,1,Seq(1,4,1,64),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,64,4,4,32,32,2,2,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,1,4,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,1,4,true,1,1,Seq(1,16,1,4),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,32,4,4,32,32,2,1,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,1,1,true,1,1,Seq(1,32,1,4),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,4,4,32,32,1,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,1,2,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,256,128,4,4,32,32,4,2,Seq(1,4,64,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,8),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,256,4,4,32,32,2,4,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,64,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,8),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,128,4,4,32,32,4,2,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,128,4,4,32,32,2,2,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,64,4,4,32,32,2,2,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,64,128,4,4,32,32,2,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,64,4,4,32,32,2,2,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,64,4,4,32,32,2,1,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,128,4,4,32,32,1,2,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,32,4,4,32,32,2,1,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,32,128,4,4,32,32,1,2,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,4,4,32,32,2,1,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,4,4,32,32,1,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,64,8,8,32,32,1,1,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,1,1,Seq(1,16,1,16),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,64,8,8,32,32,1,1,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,1,1,Seq(1,4,1,64),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,64,8,8,32,32,1,1,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,1,1,Seq(1,16,1,16),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,64,8,8,32,32,1,1,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,1,1,Seq(1,4,1,64),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp32,fp32,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp32,fp32,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp32,fp32,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<3,NDHWGC,GKZYXC,NDHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp32,fp32,0,1> diff --git a/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/nhwgc_bf16.conf b/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/nhwgc_bf16.conf index af71111410..23d72680e7 100644 --- a/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/nhwgc_bf16.conf +++ b/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/nhwgc_bf16.conf @@ -1,240 +1,240 @@ -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Default, 8, 2, 2, 2, 4, 2, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Default, 8, 2, 1, 8, 4, 2, 1, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Default, 8, 1, 2, 2, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 256, 128, 4, Default, 8, 4, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 256, 4, Default, 8, 2, 4, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 128, 4, Default, 8, 4, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 128, 4, Default, 8, 2, 2, 8, 2, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 64, 4, Default, 8, 2, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 64, 128, 4, Default, 8, 2, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Default, 8, 2, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 64, 4, Default, 8, 2, 1, 8, 2, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 128, 4, Default, 8, 1, 2, 8, 1, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Default, 8, 2, 1, 8, 4, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 32, 128, 4, Default, 8, 1, 2, 8, 1, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 32, 4, Default, 8, 2, 1, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Default, 8, 1, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Filter1x1Stride1Pad0, 8, 2, 2, 2, 4, 2, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 4, 2, 1, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Filter1x1Stride1Pad0, 8, 1, 2, 2, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 256, 128, 4, Filter1x1Stride1Pad0, 8, 4, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 256, 4, Filter1x1Stride1Pad0, 8, 2, 4, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 128, 4, Filter1x1Stride1Pad0, 8, 4, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 128, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 2, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 64, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 64, 128, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 64, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 2, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 128, 4, Filter1x1Stride1Pad0, 8, 1, 2, 8, 1, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 4, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 32, 128, 4, Filter1x1Stride1Pad0, 8, 1, 2, 8, 1, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 32, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Filter1x1Stride1Pad0, 8, 1, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<128, 16, 32, 64, Default, 8, 1, 1, 8, 8, 8, 8, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<128, 16, 64, 64, Default, 8, 1, 2, 8, 8, 8, 8, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<256, 256, 32, 64, Default, 8, 2, 1, 8, 8, 8, 8, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<256, 64, 128, 64, Default, 8, 1, 2, 8, 8, 8, 8, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<256, 64, 64, 64, Default, 8, 2, 2, 8, 8, 8, 8, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<256, 64, 64, 64, Default, 8, 2, 2, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<256, 64, 64, 64, Default, 8, 2, 2, 2, 1, 2, 1, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Filter1x1Stride1Pad0, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Filter1x1Stride1Pad0, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Filter1x1Stride1Pad0, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Filter1x1Stride1Pad0, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Filter1x1Stride1Pad0, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Filter1x1Stride1Pad0, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Filter1x1Stride1Pad0, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Filter1x1Stride1Pad0, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Filter1x1Stride1Pad0, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Filter1x1Stride1Pad0, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Filter1x1Stride1Pad0, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Filter1x1Stride1Pad0, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Filter1x1Stride1Pad0, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Filter1x1Stride1Pad0, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 8, 8, 8, 8, 1, 1, 4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 4, 8, 1, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 1, 8, 4, 8, 1, 1, 4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 2, 8, 2, 8, 1, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 1, 8, 1, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 256, 32, Default, 8, 1, 16, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 128, 32, Default, 8, 1, 8, 4, 8, 4, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 48, 64, 32, Default, 8, 3, 4, 3, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 48, 32, Default, 8, 4, 3, 4, 4, 3, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 208, 32, Default, 8, 4, 13, 4, 4, 13, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 48, 64, 32, Default, 8, 3, 4, 3, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 48, 32, Default, 8, 4, 3, 4, 4, 3, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 208, 32, Default, 8, 4, 13, 4, 4, 13, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 1x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,1,1,Seq(1,16,1,4),2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,2,1,true,1,1,Seq(1,32,1,4),2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,2,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,256,128,4,8,32,32,4,2,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,8),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,256,4,8,32,32,2,4,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,8),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,128,4,8,32,32,4,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,128,4,8,32,32,2,2,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,64,4,8,32,32,2,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,64,128,4,8,32,32,2,2,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,64,4,8,32,32,2,1,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,128,4,8,32,32,1,2,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,32,128,4,8,32,32,1,2,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,4,8,32,32,2,1,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,1,1,Seq(1,16,1,4),2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,2,1,true,1,1,Seq(1,32,1,4),2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,2,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,256,128,4,8,32,32,4,2,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,8),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,256,4,8,32,32,2,4,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,8),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,128,4,8,32,32,4,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,128,4,8,32,32,2,2,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,64,4,8,32,32,2,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,64,128,4,8,32,32,2,2,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,64,4,8,32,32,2,1,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,128,4,8,32,32,1,2,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,32,128,4,8,32,32,1,2,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,4,8,32,32,2,1,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,16,32,64,8,16,16,1,1,Seq(8,2,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(4,4,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,16,1,8),4,Intrawave,v1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,128,16,64,64,8,16,16,1,2,Seq(8,2,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(2,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,16,1,8),4,Intrawave,v1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,256,32,64,8,32,32,2,1,Seq(4,32,2),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(8,4,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,64,1,4),8,Intrawave,v1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,128,64,8,32,32,1,2,Seq(4,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(4,16,4),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,32,1,8),8,Intrawave,v1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,16,16,2,2,Seq(4,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(4,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,32,1,4),8,Intrawave,v1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,16,16,2,2,Seq(4,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(4,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,32,1,4),2,Intrawave,v1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,16,16,2,2,Seq(4,32,2),Seq(0,2,1),Seq(0,2,1),1,2,1,false,Seq(4,32,2),Seq(0,2,1),Seq(0,2,1),1,2,1,false,1,1,Seq(1,32,1,4),4,Intrawave,v1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,bf16,bf16,0,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,4,1),Seq(2,0,1),Seq(2,0,1),1,8,8,false,Seq(8,4,1),Seq(2,0,1),Seq(2,0,1),1,8,8,false,1,1,Seq(1,16,1,16),4,Intrawave,v1,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,16,1),Seq(2,0,1),Seq(2,0,1),1,4,8,false,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,1,1,Seq(1,4,1,64),1,Intrawave,v1,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,Seq(8,16,1),Seq(2,0,1),Seq(2,0,1),1,4,8,false,1,1,Seq(1,16,1,16),4,Intrawave,v1,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,2,8,false,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,2,8,false,1,1,Seq(1,8,1,32),2,Intrawave,v1,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,1,1,Seq(1,4,1,64),1,Intrawave,v1,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,256,32,8,16,16,1,16,Seq(4,2,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,4,1,16),1,Intrawave,v1,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,128,32,8,16,16,1,8,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,4,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,8,false,1,1,Seq(1,4,1,16),1,Intrawave,v1,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,2,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,4,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,8,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,48,64,32,8,16,16,3,4,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,48,32,8,16,16,4,3,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,208,32,8,16,16,4,13,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,13,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,48,64,32,8,16,16,3,4,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,48,32,8,16,16,4,3,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,bf16,bf16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,bf16,bf16,bf16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,208,32,8,16,16,4,13,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,13,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,bf16,bf16,1,1> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/nhwgc_fp16.conf b/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/nhwgc_fp16.conf index ae895fde7e..f2d83b4c0c 100644 --- a/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/nhwgc_fp16.conf +++ b/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/nhwgc_fp16.conf @@ -1,243 +1,243 @@ -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Default, 8, 2, 2, 2, 4, 2, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Default, 8, 2, 1, 8, 4, 2, 1, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Default, 8, 1, 2, 2, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 256, 128, 4, Default, 8, 4, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 256, 4, Default, 8, 2, 4, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 128, 4, Default, 8, 4, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 128, 4, Default, 8, 2, 2, 8, 2, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 64, 4, Default, 8, 2, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 64, 128, 4, Default, 8, 2, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Default, 8, 2, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 64, 4, Default, 8, 2, 1, 8, 2, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 128, 4, Default, 8, 1, 2, 8, 1, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Default, 8, 2, 1, 8, 4, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 32, 128, 4, Default, 8, 1, 2, 8, 1, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 32, 4, Default, 8, 2, 1, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Default, 8, 1, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Filter1x1Stride1Pad0, 8, 2, 2, 2, 4, 2, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 4, 2, 1, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Filter1x1Stride1Pad0, 8, 1, 2, 2, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 256, 128, 4, Filter1x1Stride1Pad0, 8, 4, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 256, 4, Filter1x1Stride1Pad0, 8, 2, 4, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 128, 4, Filter1x1Stride1Pad0, 8, 4, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 128, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 2, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 64, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 64, 128, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Filter1x1Stride1Pad0, 8, 2, 2, 8, 4, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 64, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 2, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 128, 4, Filter1x1Stride1Pad0, 8, 1, 2, 8, 1, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 4, 8, 1, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 32, 128, 4, Filter1x1Stride1Pad0, 8, 1, 2, 8, 1, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 32, 4, Filter1x1Stride1Pad0, 8, 2, 1, 8, 4, 8, 2, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Filter1x1Stride1Pad0, 8, 1, 2, 8, 2, 8, 4, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<128, 16, 32, 64, Default, 8, 1, 1, 8, 8, 8, 8, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<128, 16, 32, 64, Default, 8, 1, 1, 8, 8, 8, 8, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<128, 16, 64, 64, Default, 8, 1, 2, 8, 8, 8, 8, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<128, 16, 64, 64, Default, 8, 1, 2, 8, 8, 8, 8, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<256, 256, 32, 64, Default, 8, 2, 1, 8, 8, 8, 8, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<256, 64, 128, 64, Default, 8, 1, 2, 8, 8, 8, 8, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<256, 64, 128, 64, Default, 8, 1, 2, 8, 8, 8, 8, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<256, 64, 64, 64, Default, 8, 2, 2, 8, 8, 8, 8, 1, 1, 8> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<256, 64, 64, 64, Default, 8, 2, 2, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3_DirectLoad<256, 64, 64, 64, Default, 8, 2, 2, 2, 1, 2, 1, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Filter1x1Stride1Pad0, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Filter1x1Stride1Pad0, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Filter1x1Stride1Pad0, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Filter1x1Stride1Pad0, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Filter1x1Stride1Pad0, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Filter1x1Stride1Pad0, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Filter1x1Stride1Pad0, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 32, 32, Filter1x1Stride1Pad0, 8, 1, 1, 2, 2, 2, 2, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 64, 32, Filter1x1Stride1Pad0, 8, 1, 2, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 32, 128, 32, Filter1x1Stride1Pad0, 8, 1, 4, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 32, 32, Filter1x1Stride1Pad0, 8, 2, 1, 4, 4, 4, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Filter1x1Stride1Pad0, 8, 4, 1, 8, 8, 8, 8, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 80, 32, Filter1x1Stride1Pad0, 8, 4, 5, 4, 4, 5, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 64, 112, 32, Filter1x1Stride1Pad0, 8, 4, 7, 4, 4, 7, 4, 1, 1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 8, 8, 8, 8, 1, 1, 4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 4, 8, 1, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 1, 8, 4, 8, 1, 1, 4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 2, 8, 2, 8, 1, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<256, 64, 64, 64, Default, 8, 1, 1, 1, 8, 1, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 256, 32, Default, 8, 1, 16, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 128, 32, Default, 8, 1, 8, 4, 8, 4, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 64, 32, Default, 8, 1, 2, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 128, 32, Default, 8, 1, 4, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 32, 32, 32, Default, 8, 1, 1, 2, 2, 2, 2, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 2> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 32, 32, Default, 8, 2, 1, 4, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 4> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 128, 32, 32, Default, 8, 4, 1, 8, 8, 8, 8, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 8> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 48, 64, 32, Default, 8, 3, 4, 3, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 48, 32, Default, 8, 4, 3, 4, 4, 3, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 208, 32, Default, 8, 4, 13, 4, 4, 13, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 48, 64, 32, Default, 8, 3, 4, 3, 4, 4, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 48, 32, Default, 8, 4, 3, 4, 4, 3, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 80, 32, Default, 8, 4, 5, 4, 4, 5, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 112, 32, Default, 8, 4, 7, 4, 4, 7, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<64, 64, 208, 32, Default, 8, 4, 13, 4, 4, 13, 4, 1, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x256x32, WaveTile: 32x32, WaveMap: 4x4, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 128x128x64, WaveTile: 32x32, WaveMap: 2x2, VmemReadVec: 8x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x4, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x8, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x2, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 8x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 8x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 4x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 2x1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1, BlkGemmPipelinePrefetchStages: 1> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 256x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 128x16x64, WaveTile: 16x16, WaveMap: 4x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 64x16x64, WaveTile: 16x16, WaveMap: 2x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 32x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 64, BlkTile: 16x16x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x32x64, WaveTile: 16x16, WaveMap: 1x1, VmemReadVec: 1x4, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 3> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x64x64, WaveTile: 16x16, WaveMap: 1x2, VmemReadVec: 1x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 128, BlkTile: 16x128x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x8, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> -DeviceGroupedConvBwdWeight_Explicit_Xdl BlkSize: 256, BlkTile: 16x256x64, WaveTile: 16x16, WaveMap: 1x4, VmemReadVec: 1x2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2, BlkGemmPipelinePrefetchStages: 2> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,1,1,Seq(1,16,1,4),2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,2,1,true,1,1,Seq(1,32,1,4),2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,2,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,256,128,4,8,32,32,4,2,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,8),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,256,4,8,32,32,2,4,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,8),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,128,4,8,32,32,4,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,128,4,8,32,32,2,2,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,64,4,8,32,32,2,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,64,128,4,8,32,32,2,2,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,64,4,8,32,32,2,1,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,128,4,8,32,32,1,2,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,32,128,4,8,32,32,1,2,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,4,8,32,32,2,1,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,2,4,true,1,1,Seq(1,16,1,4),2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,2,1,true,1,1,Seq(1,32,1,4),2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,2,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,256,128,4,8,32,32,4,2,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,8),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,256,4,8,32,32,2,4,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,8),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,128,4,8,32,32,4,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,128,4,8,32,32,2,2,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,64,4,8,32,32,2,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,64,128,4,8,32,32,2,2,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,64,4,8,32,32,2,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,64,4,8,32,32,2,1,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,128,4,8,32,32,1,2,Seq(1,4,8,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,32,4,8,32,32,2,1,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,32,128,4,8,32,32,1,2,Seq(1,4,4,8),Seq(0,3,1,2),Seq(0,2,1,3),2,8,1,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,32,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,4,8,32,32,2,1,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,4,8,32,32,1,2,Seq(1,4,4,4),Seq(0,3,1,2),Seq(0,2,1,3),2,8,2,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,8,4,true,1,1,Seq(1,16,1,4),8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,16,32,64,8,16,16,1,1,Seq(8,2,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(4,4,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,16,1,8),4,Intrawave,v1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,16,32,64,8,16,16,1,1,Seq(8,2,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(4,4,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,16,1,8),4,Intrawave,v1,fp16,fp16,1,2> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,16,64,64,8,16,16,1,2,Seq(8,2,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(2,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,16,1,8),4,Intrawave,v1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,128,16,64,64,8,16,16,1,2,Seq(8,2,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(2,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,16,1,8),4,Intrawave,v1,fp16,fp16,1,2> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,256,32,64,8,32,32,2,1,Seq(4,32,2),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(8,4,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,64,1,4),8,Intrawave,v1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,128,64,8,32,32,1,2,Seq(4,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(4,16,4),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,32,1,8),8,Intrawave,v1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,128,64,8,32,32,1,2,Seq(4,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(4,16,4),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,32,1,8),8,Intrawave,v1,fp16,fp16,1,2> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,16,16,2,2,Seq(4,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(4,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,32,1,4),8,Intrawave,v1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,16,16,2,2,Seq(4,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,Seq(4,8,8),Seq(0,2,1),Seq(0,2,1),1,8,8,false,1,1,Seq(1,32,1,4),2,Intrawave,v1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,16,16,2,2,Seq(4,32,2),Seq(0,2,1),Seq(0,2,1),1,2,1,false,Seq(4,32,2),Seq(0,2,1),Seq(0,2,1),1,2,1,false,1,1,Seq(1,32,1,4),4,Intrawave,v1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp16,fp16,0,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v1,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,8,1),Seq(2,0,1),Seq(2,0,1),1,8,8,false,Seq(8,8,1),Seq(2,0,1),Seq(2,0,1),1,8,8,false,1,1,Seq(1,16,1,16),4,Intrawave,v1,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,16,1),Seq(2,0,1),Seq(2,0,1),1,4,8,false,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,1,1,Seq(1,4,1,64),1,Intrawave,v1,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,Seq(8,16,1),Seq(2,0,1),Seq(2,0,1),1,4,8,false,1,1,Seq(1,16,1,16),4,Intrawave,v1,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,2,8,false,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,2,8,false,1,1,Seq(1,8,1,32),2,Intrawave,v1,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,64,8,32,32,1,1,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,Seq(8,32,1),Seq(2,0,1),Seq(2,0,1),1,1,8,false,1,1,Seq(1,4,1,64),1,Intrawave,v1,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,256,32,8,16,16,1,16,Seq(4,2,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,4,1,16),1,Intrawave,v1,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,128,32,8,16,16,1,8,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,4,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,8,false,1,1,Seq(1,4,1,16),1,Intrawave,v1,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,32,8,32,32,1,2,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,128,32,8,32,32,1,4,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,32,32,8,32,32,1,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,2,2,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,2,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,32,8,32,32,2,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,4,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,128,32,32,8,32,32,4,1,Seq(4,16,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,Seq(4,4,1),Seq(2,0,1),Seq(1,0,2),1,8,8,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,8,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,48,64,32,8,16,16,3,4,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,48,32,8,16,16,4,3,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,208,32,8,16,16,4,13,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,13,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v2,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,48,64,32,8,16,16,3,4,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,48,32,8,16,16,4,3,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,3,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,80,32,8,16,16,4,5,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,5,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,112,32,8,16,16,4,7,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,7,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,fp16,fp16,1,1> +DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp16,fp16,fp16,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,208,32,8,16,16,4,13,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,4,4,false,Seq(4,16,1),Seq(2,0,1),Seq(2,0,1),1,13,4,false,1,1,Seq(1,8,1,8),1,Intrawave,v5,1,fp16,fp16,1,1> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/nhwgc_fp32.conf b/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/nhwgc_fp32.conf index b6386d41ad..e9ac3d1072 100644 --- a/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/nhwgc_fp32.conf +++ b/experimental/grouped_convolution_tile_instances/configs/bwd_weight/profiler/nhwgc_fp32.conf @@ -1,44 +1,44 @@ -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Default, 4, 2, 2, 1, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Default, 4, 2, 1, 4, 4, 1, 1, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Default, 4, 1, 2, 1, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 256, 128, 4, Default, 4, 4, 2, 4, 4, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 256, 4, Default, 4, 2, 4, 4, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 128, 4, Default, 4, 4, 2, 4, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 128, 4, Default, 4, 2, 2, 4, 2, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 64, 4, Default, 4, 2, 2, 4, 4, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 64, 128, 4, Default, 4, 2, 2, 4, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Default, 4, 2, 2, 4, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 64, 4, Default, 4, 2, 1, 4, 2, 4, 1, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 128, 4, Default, 4, 1, 2, 4, 1, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Default, 4, 2, 1, 4, 4, 4, 1, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 32, 128, 4, Default, 4, 1, 2, 4, 1, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 32, 4, Default, 4, 2, 1, 4, 4, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Default, 4, 1, 2, 4, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Default, 8, 1, 1, 4, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Default, 8, 1, 1, 4, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Default, 8, 1, 1, 1, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Filter1x1Stride1Pad0, 4, 2, 2, 1, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Filter1x1Stride1Pad0, 4, 2, 1, 4, 4, 1, 1, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Filter1x1Stride1Pad0, 4, 1, 2, 1, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 256, 128, 4, Filter1x1Stride1Pad0, 4, 4, 2, 4, 4, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 256, 4, Filter1x1Stride1Pad0, 4, 2, 4, 4, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 128, 4, Filter1x1Stride1Pad0, 4, 4, 2, 4, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 128, 4, Filter1x1Stride1Pad0, 4, 2, 2, 4, 2, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 64, 4, Filter1x1Stride1Pad0, 4, 2, 2, 4, 4, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 64, 128, 4, Filter1x1Stride1Pad0, 4, 2, 2, 4, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 64, 4, Filter1x1Stride1Pad0, 4, 2, 2, 4, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 128, 64, 4, Filter1x1Stride1Pad0, 4, 2, 1, 4, 2, 4, 1, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 128, 4, Filter1x1Stride1Pad0, 4, 1, 2, 4, 1, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 128, 32, 4, Filter1x1Stride1Pad0, 4, 2, 1, 4, 4, 4, 1, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<128, 32, 128, 4, Filter1x1Stride1Pad0, 4, 1, 2, 4, 1, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 64, 32, 4, Filter1x1Stride1Pad0, 4, 2, 1, 4, 4, 4, 2, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<64, 32, 64, 4, Filter1x1Stride1Pad0, 4, 1, 2, 4, 2, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Filter1x1Stride1Pad0, 8, 1, 1, 4, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Filter1x1Stride1Pad0, 8, 1, 1, 4, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Filter1x1Stride1Pad0, 8, 1, 1, 1, 4, 4, 4, 1, 1, 4> -DeviceGroupedConvBwdWeight_Xdl_CShuffle<256, 64, 64, 8, Filter1x1Stride1Pad0, 8, 1, 1, 1, 4, 1, 4, 1, 1, 1> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 16, 16, 32, Default, 8, 1, 1, 1, 4, 1, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 16, 16, 32, Filter1x1Stride1Pad0, 8, 1, 1, 1, 4, 1, 4, 1, 1, 2> -DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 16, 16, 32, Filter1x1Stride1Pad0, 8, 1, 1, 1, 4, 1, 4, 1, 1, 2> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,64,4,4,32,32,2,2,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,1,4,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,1,4,true,1,1,Seq(1,16,1,4),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,32,4,4,32,32,2,1,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,1,1,true,1,1,Seq(1,32,1,4),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,4,4,32,32,1,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,1,2,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,256,128,4,4,32,32,4,2,Seq(1,4,64,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,8),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,256,4,4,32,32,2,4,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,64,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,8),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,128,4,4,32,32,4,2,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,128,4,4,32,32,2,2,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,64,4,4,32,32,2,2,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,128,64,128,4,4,32,32,2,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,64,4,4,32,32,2,2,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,128,64,4,4,32,32,2,1,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,128,4,4,32,32,1,2,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,128,128,32,4,4,32,32,2,1,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,128,32,128,4,4,32,32,1,2,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,64,32,4,4,32,32,2,1,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,32,64,4,4,32,32,1,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,8,8,32,32,1,1,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,1,1,Seq(1,16,1,16),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,8,8,32,32,1,1,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,1,1,Seq(1,4,1,64),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,8,8,32,32,1,1,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,1,1,Seq(1,16,1,16),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,256,64,64,8,8,32,32,1,1,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,1,1,Seq(1,4,1,64),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,64,4,4,32,32,2,2,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,1,4,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,1,4,true,1,1,Seq(1,16,1,4),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,32,4,4,32,32,2,1,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,1,1,true,1,1,Seq(1,32,1,4),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,4,4,32,32,1,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,1,2,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,256,128,4,4,32,32,4,2,Seq(1,4,64,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,8),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,256,4,4,32,32,2,4,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,64,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,8),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,128,4,4,32,32,4,2,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,128,4,4,32,32,2,2,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,64,4,4,32,32,2,2,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,64,128,4,4,32,32,2,2,Seq(1,4,16,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,64,4,4,32,32,2,2,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,128,64,4,4,32,32,2,1,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,128,4,4,32,32,1,2,Seq(1,4,16,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,Seq(1,4,32,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,128,32,4,4,32,32,2,1,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,128,32,128,4,4,32,32,1,2,Seq(1,4,8,4),Seq(0,3,1,2),Seq(0,2,1,3),2,4,1,true,Seq(1,4,32,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,32,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,64,32,4,4,32,32,2,1,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,32,64,4,4,32,32,1,2,Seq(1,4,8,2),Seq(0,3,1,2),Seq(0,2,1,3),2,4,2,true,Seq(1,4,16,1),Seq(0,3,1,2),Seq(0,2,1,3),2,4,4,true,1,1,Seq(1,16,1,4),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,64,8,8,32,32,1,1,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,1,1,Seq(1,16,1,16),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,64,8,8,32,32,1,1,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,1,1,Seq(1,4,1,64),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,64,8,8,32,32,1,1,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,Seq(1,8,16,1),Seq(0,3,1,2),Seq(0,3,1,2),2,4,4,true,1,1,Seq(1,16,1,16),4,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffle<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,256,64,64,8,8,32,32,1,1,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,Seq(1,8,32,1),Seq(0,3,1,2),Seq(0,3,1,2),2,1,4,true,1,1,Seq(1,4,1,64),1,fp32,fp32,1,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp32,fp32,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Default,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp32,fp32,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v2,fp32,fp32,0,1> +DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<2,NHWGC,GKYXC,NHWGK,fp32,fp32,fp32,fp32,PassThrough,PassThrough,PassThrough,Filter1x1Stride1Pad0,64,16,16,32,8,16,16,1,1,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,Seq(4,8,1),Seq(2,0,1),Seq(1,0,2),1,1,4,false,1,1,Seq(1,8,1,8),2,Intrawave,v5,fp32,fp32,0,1>