ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-07-02 21:27:45 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
c3228aaf0dfc4160142ef4444a5ba1dfd195f081
composable_kernel/include/ck/tensor_operation/gpu/device/impl
History
Qun Lin c3228aaf0d [ck] support remap 32x32 warp tile to 16x16 
This PR remap 32x32 warp tile to 16x16 warp tile for all CK kernels in wave32. the logic is same with ROCm/composable_kernel#3421. and the most change is in device classes.

To reduece the instance build time, VGPR estimation is implemented in ~10 gridwise classes. and to pass all test in CI, several tests are minor adjusted.
2026-01-12 14:30:09 +08:00
..
codegen_device_grouped_conv_fwd_multiple_abd_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_avgpool2d_bwd_nhwc_nhwc.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_avgpool3d_bwd_ndhwc_ndhwc.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_batched_contraction_multiple_d_wmma_cshuffle.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_batched_contraction_multiple_d_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_batched_gemm_e_permute_xdl.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_batched_gemm_gemm_wmma_cshuffle_v3.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_batched_gemm_gemm_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_batched_gemm_multi_d_xdl.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_batched_gemm_multiple_d_dl.hpp
Add support for RDNA1 GPUs (#3220)
2025-11-20 10:45:57 -08:00
device_batched_gemm_multiple_d_gemm_multiple_d_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_batched_gemm_multiple_d_wmma_cshuffle_v3.hpp
Wmma support for grouped convolution bwd weight (#2947)
2025-12-17 15:58:58 -08:00
device_batched_gemm_multiple_d_xdl_cshuffle_v3.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_batched_gemm_reduce_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_batched_gemm_softmax_gemm_permute_wmma_cshuffle.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_batched_gemm_softmax_gemm_permute_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_batched_gemm_softmax_gemm_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_batched_gemm_wmma_cshuffle_v3_b_scale.hpp
Wmma support for gemm_ab_scale (#3314)
2025-12-11 09:06:20 +01:00
device_batched_gemm_wmma_cshuffle_v3.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_batched_gemm_xdl_fpAintB_b_scale.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_batched_gemm_xdl.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_batchnorm_backward_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_batchnorm_forward_impl_obsolete.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_batchnorm_forward_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_cgemm_4gemm_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_column_to_image_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_contraction_multiple_abd_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_contraction_multiple_d_xdl_cshuffle.hpp
[ck] add gridwise base class for in all xdl kernel (#186)
2026-01-12 10:01:53 +08:00
device_contraction_utils.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_conv2d_backward_weight_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_conv2d_fwd_xdl_c_shuffle_bias_activation_add_nhwc_kyxc_nhwk.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_conv2d_fwd_xdl_c_shuffle_bias_activation_nhwc_kyxc_nhwk.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_conv2d_fwd_xdl_nhwc_kyxc_nhwk.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_conv3d_fwd_naive_ndhwc_kzyxc_ndhwk.hpp
[CK] Integrate GPU reference into ckProfiler for convolutions (#3379)
2025-12-18 07:59:45 +01:00
device_conv3d_fwd_xdl_ndhwc_kzyxc_ndhwk.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_convnd_bwd_data_nwc_kxc_nwk_dl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_convnd_bwd_data_nwc_kxc_nwk_xdl.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_elementwise_dynamic_vector_dims_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_elementwise_normalization_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_elementwise_scale_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_fpAintB_gemm_wmma.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_gemm_bias_add_reduce_wmma_cshuffle_v3.hpp
Wmma support for gemm_bias_add_reduce (#3316)
2026-01-07 10:27:16 -08:00
device_gemm_bias_add_reduce_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_dl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_gemm_dpp.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_gemm_multiple_abd_wmma_cshuffle_v3.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_gemm_multiple_abd_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_multiple_d_dl.hpp
Add support for RDNA1 GPUs (#3220)
2025-11-20 10:45:57 -08:00
device_gemm_multiple_d_layernorm_wmma_cshuffle_v3.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_gemm_multiple_d_layernorm_xdl_cshuffle.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_gemm_multiple_d_multiple_r_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_multiple_d_wmma_cshuffle_v3_ab_scale.hpp
Wmma support for gemm_ab_scale (#3314)
2025-12-11 09:06:20 +01:00
device_gemm_multiple_d_wmma_cshuffle_v3_b_preshuffle.hpp
Wmma support for gemm_ab_scale (#3314)
2025-12-11 09:06:20 +01:00
device_gemm_multiple_d_wmma_cshuffle_v3_blockscale_bpreshuffle.hpp
Wmma support for gemm_ab_scale (#3314)
2025-12-11 09:06:20 +01:00
device_gemm_multiple_d_wmma_cshuffle_v3.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_gemm_multiple_d_wmma_cshuffle.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_gemm_multiple_d_xdl_cshuffle_lds_direct_load.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_multiple_d_xdl_cshuffle_v3_ab_scale.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_multiple_d_xdl_cshuffle_v3_b_preshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_multiple_d_xdl_cshuffle_v3_blockscale_bpreshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_multiple_d_xdl_cshuffle_v3.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_multiple_d_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_reduce_wmma_cshuffle_v3.hpp
Wmma support for gemm_bias_add_reduce (#3316)
2026-01-07 10:27:16 -08:00
device_gemm_reduce_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_wmma_cshuffle_v3_b_scale.hpp
Wmma support for gemm_ab_scale (#3314)
2025-12-11 09:06:20 +01:00
device_gemm_wmma_cshuffle_v3_common.hpp
Wmma support for gemm_ab_scale (#3314)
2025-12-11 09:06:20 +01:00
device_gemm_wmma_cshuffle_v3.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_gemm_wmma_cshuffle_v3r1.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_gemm_wmma.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_gemm_xdl_cshuffle_lds_direct_load.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_cshuffle_streamk_v3.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_cshuffle_v2.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_cshuffle_v3_b_preshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_cshuffle_v3_b_scale.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_cshuffle_v3_mx.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_cshuffle_v3.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_cshuffle_v3r1.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_layernorm_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_skip_b_lds.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_splitk_c_shuffle_lds_direct_load.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_splitk_c_shuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_streamk.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl_waveletmodel_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_gemm_xdl.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_contraction_multiple_d_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_conv_bwd_data_multiple_d_wmma_cshuffle_v3.hpp
Disable fp32 atomic adds on gfx11 (#3510)
2026-01-07 15:32:04 -08:00
device_grouped_conv_bwd_data_multiple_d_wmma_cshuffle.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_grouped_conv_bwd_data_multiple_d_xdl_cshuffle_v1.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_conv_bwd_weight_dl.hpp
[CK_BUILDER] Instance traits for conv bwd weight algorithms (#3498)
2025-12-31 15:41:15 -08:00
device_grouped_conv_bwd_weight_explicit.hpp
Wmma support for grouped convolution bwd weight (#2947)
2025-12-17 15:58:58 -08:00
device_grouped_conv_bwd_weight_multiple_d_wmma_cshuffle_v3.hpp
Disable fp32 atomic adds on gfx11 (#3510)
2026-01-07 15:32:04 -08:00
device_grouped_conv_bwd_weight_multiple_d_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_conv_bwd_weight_two_stage_wmma_cshuffle_v3.hpp
Disable fp32 atomic adds on gfx11 (#3510)
2026-01-07 15:32:04 -08:00
device_grouped_conv_bwd_weight_two_stage_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_conv_bwd_weight_wmma_cshuffle_v3.hpp
Disable fp32 atomic adds on gfx11 (#3510)
2026-01-07 15:32:04 -08:00
device_grouped_conv_bwd_weight_wmma_cshuffle.hpp
[CK_BUILDER] Instance traits for conv bwd weight algorithms (#3498)
2025-12-31 15:41:15 -08:00
device_grouped_conv_bwd_weight_xdl_cshuffle_v3.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_conv_bwd_weight_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_conv_fwd_dl_multiple_d_nhwc_kyxc_nhwk.hpp
Add describe() method to device ops for runtime introspection (#3375)
2025-12-14 12:49:12 -08:00
device_grouped_conv_fwd_dl_nhwc_kyxc_nhwk.hpp
Add support for RDNA1 GPUs (#3220)
2025-11-20 10:45:57 -08:00
device_grouped_conv_fwd_multiple_abd_wmma_cshuffle_v3.hpp
Grouped convolution forward device implementation and base flavors for RDNA3/4 (#2964)
2025-12-18 13:12:15 -07:00
device_grouped_conv_fwd_multiple_abd_xdl_cshuffle_v3.hpp
[ck] add gridwise base class for in all xdl kernel (#186)
2026-01-12 10:01:53 +08:00
device_grouped_conv_fwd_multiple_abd_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_conv_fwd_multiple_d_multiple_r_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_conv_fwd_multiple_d_multiple_r.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_grouped_conv_fwd_multiple_d_wmma_cshuffle_v3_large_tensor.hpp
Added large tensor support for grouped conv fwd wmma (#3437)
2025-12-18 21:55:50 -07:00
device_grouped_conv_fwd_multiple_d_wmma_cshuffle.hpp
Add describe() method to device ops for runtime introspection (#3375)
2025-12-14 12:49:12 -08:00
device_grouped_conv_fwd_multiple_d_xdl_cshuffle.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_grouped_conv_fwd_multiple_d_xdl_large_tensor_cshuffle.hpp
[ck] add gridwise base class for in all xdl kernel (#186)
2026-01-12 10:01:53 +08:00
device_grouped_conv_utils.hpp
Grouped convolution forward device implementation and base flavors for RDNA3/4 (#2964)
2025-12-18 13:12:15 -07:00
device_grouped_gemm_multi_abd_xdl_fixed_nk.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_gemm_multiple_d_dl.hpp
Add support for RDNA1 GPUs (#3220)
2025-11-20 10:45:57 -08:00
device_grouped_gemm_multiple_d_splitk_xdl_cshuffle_two_stage.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_gemm_multiple_d_xdl_cshuffle_tile_loop.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_gemm_softmax_gemm_permute_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_gemm_wmma_splitk_cshuffle_v3.hpp
Implement grouped gemm fastgelu for RDNA4 (#3303)
2026-01-07 10:20:44 -08:00
device_grouped_gemm_xdl_fixed_nk.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_gemm_xdl_splitk_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_gemm_xdl.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_grouped_query_attention_forward_wmma.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_image_to_column_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_max_pool_bwd_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_moe_gemm_blockscale.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_moe_gemm.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_moe_mx_gemm_bns.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_moe_mx_gemm_bpreshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_moe_mx_gemm.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
device_multi_query_attention_forward_wmma.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_multiple_reduce_multiblock.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_multiple_reduce_threadwise.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_normalization_bwd_data_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_normalization_bwd_gamma_beta_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_normalization_fwd_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_normalization_fwd_splitk_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_permute_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_pool2d_fwd_nhwc_nhwc.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_pool3d_fwd_ndhwc_ndhwc.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_put_element_impl.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_reduce_common.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_reduce_multiblock.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_reduce_threadwise_multi_d.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_reduce_threadwise.hpp
chore(copyright): update copyright header for include directory (#3224)
2025-11-18 10:17:18 -08:00
device_softmax_impl.hpp
chore(copyright): update copyright header for include directory (#3293)
2025-11-26 11:00:05 -07:00
device_sparse_embeddings_forward_layernorm.hpp
chore(copyright): update copyright header for include directory (#3293)
2025-11-26 11:00:05 -07:00
device_splitk_contraction_multiple_d_xdl_cshuffle.hpp
[ck] support remap 32x32 warp tile to 16x16
2026-01-12 14:30:09 +08:00
split_k_arg.hpp
chore(copyright): update copyright header for include directory (#3293)
2025-11-26 11:00:05 -07:00
split_k_offset_utils.hpp
[CK] Allow tensors larger than 2GB in grouped conv bwd weight (#3169)
2026-01-08 08:02:02 +01:00
split_k_utils.hpp
chore(copyright): update copyright header for include directory (#3293)
2025-11-26 11:00:05 -07:00