ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-03 13:11:25 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
507d81c3af51b81f15b946a2a4bef7f594620292
composable_kernel/include/ck/tensor_operation/gpu/device/impl
History
Enrico Degregori 507d81c3af Fix splitk preshuffle (#3137) 
* Fix splitK multiply_multiply_wp

* Add tests for gemm_multiply_multiply_wp

* Add tests for gemm_universal_preshuffle (KBatch = 1)

* Add tests gemm_blockscale_wp

* Fix splitk gemm universal preshuffle

* Run new tests on arch supporting fp8

* Restore example

* Fix strides profiler

* Fix tests

* Fix clang format

* Finalize profiler preshuffle with tolerances

* Minor improvements to splitk related changes

* Address review comments: clang format and ckProfiler typo

* Remove b_k_split_offset from SplitKBatchOffset struct
2025-11-03 11:59:01 -08:00
..
codegen_device_grouped_conv_fwd_multiple_abd_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_avgpool2d_bwd_nhwc_nhwc.hpp
Pool2d max/avg kernel in the BWD version (#1494)
2024-09-12 11:47:52 +02:00
device_avgpool3d_bwd_ndhwc_ndhwc.hpp
Average pool backward deviceOP and example (#797)
2023-08-10 12:04:35 +08:00
device_batched_contraction_multiple_d_wmma_cshuffle.hpp
Disable GridwiseOp prints if env var is off (#2843)
2025-09-16 17:47:28 +02:00
device_batched_contraction_multiple_d_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_batched_gemm_e_permute_xdl.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_batched_gemm_gemm_wmma_cshuffle_v3.hpp
Implement batched gemm gemm for RDNA (3 and 4) (#2612)
2025-09-04 14:10:24 -07:00
device_batched_gemm_gemm_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_batched_gemm_multi_d_xdl.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_batched_gemm_multiple_d_dl.hpp
Remove !defined(__HIP_DEVICE_COMPILE__) in CK kernel (#2564)
2025-07-28 13:01:07 -07:00
device_batched_gemm_multiple_d_gemm_multiple_d_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_batched_gemm_multiple_d_xdl_cshuffle_v3.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_batched_gemm_reduce_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_batched_gemm_softmax_gemm_permute_wmma_cshuffle.hpp
Remove !defined(__HIP_DEVICE_COMPILE__) in CK kernel (#2564)
2025-07-28 13:01:07 -07:00
device_batched_gemm_softmax_gemm_permute_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_batched_gemm_softmax_gemm_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_batched_gemm_wmma_cshuffle_v3_b_scale.hpp
WMMA gemm_add_relu_add_layernorm (#2989)
2025-10-31 11:19:26 -07:00
device_batched_gemm_wmma_cshuffle_v3.hpp
WMMA gemm_add_relu_add_layernorm (#2989)
2025-10-31 11:19:26 -07:00
device_batched_gemm_xdl_fpAintB_b_scale.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_batched_gemm_xdl.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_batchnorm_backward_impl.hpp
Fixed GroupedGemmFixedNK with hipGraph (#1065)
2023-11-30 15:09:27 -06:00
device_batchnorm_forward_impl_obsolete.hpp
Fixed GroupedGemmFixedNK with hipGraph (#1065)
2023-11-30 15:09:27 -06:00
device_batchnorm_forward_impl.hpp
Fixed GroupedGemmFixedNK with hipGraph (#1065)
2023-11-30 15:09:27 -06:00
device_cgemm_4gemm_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_column_to_image_impl.hpp
upgrade from clang-format-12 to clang-format-18 (#2568)
2025-07-28 11:34:07 -07:00
device_contraction_multiple_abd_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_contraction_multiple_d_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_contraction_utils.hpp
upgrade from clang-format-12 to clang-format-18 (#2568)
2025-07-28 11:34:07 -07:00
device_conv2d_backward_weight_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
Fixed handling of split-K autodeduce argument for grouped convolution (#3024)
2025-10-17 15:36:39 +03:00
device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_conv2d_fwd_xdl_c_shuffle_bias_activation_add_nhwc_kyxc_nhwk.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_conv2d_fwd_xdl_c_shuffle_bias_activation_nhwc_kyxc_nhwk.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_conv2d_fwd_xdl_nhwc_kyxc_nhwk.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_conv3d_fwd_naive_ndhwc_kzyxc_ndhwk.hpp
update copyright headers (#726)
2023-05-31 18:46:57 -05:00
device_conv3d_fwd_xdl_ndhwc_kzyxc_ndhwk.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_convnd_bwd_data_nwc_kxc_nwk_dl.hpp
Split env.hpp header from the ck.hpp header. (#2049)
2025-04-03 15:30:21 -07:00
device_convnd_bwd_data_nwc_kxc_nwk_xdl.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_elementwise_dynamic_vector_dims_impl.hpp
Refactor elementwise kernels (#1222)
2024-04-19 13:31:17 +02:00
device_elementwise_normalization_impl.hpp
update copyright headers (#726)
2023-05-31 18:46:57 -05:00
device_elementwise_scale_impl.hpp
Refactor elementwise kernels (#1222)
2024-04-19 13:31:17 +02:00
device_fpAintB_gemm_wmma.hpp
Merging the gfx12 code into public repo. (#1362)
2024-06-27 00:33:34 -07:00
device_gemm_bias_add_reduce_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_dl.hpp
Split env.hpp header from the ck.hpp header. (#2049)
2025-04-03 15:30:21 -07:00
device_gemm_dpp.hpp
Code clean-up (#1285)
2024-05-10 09:41:39 -07:00
device_gemm_multiple_abd_wmma_cshuffle_v3.hpp
Wmma support for multiple ABD GEMM (#2803)
2025-09-22 18:49:06 -07:00
device_gemm_multiple_abd_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_multiple_d_dl.hpp
Remove !defined(__HIP_DEVICE_COMPILE__) in CK kernel (#2564)
2025-07-28 13:01:07 -07:00
device_gemm_multiple_d_layernorm_wmma_cshuffle_v3.hpp
WMMA gemm_add_relu_add_layernorm (#2989)
2025-10-31 11:19:26 -07:00
device_gemm_multiple_d_layernorm_xdl_cshuffle.hpp
[CK] Fix misc issues in CK examples (#2890)
2025-09-24 11:28:20 -07:00
device_gemm_multiple_d_multiple_r_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_multiple_d_wmma_cshuffle_v3.hpp
Wmma support for multiple ABD GEMM (#2803)
2025-09-22 18:49:06 -07:00
device_gemm_multiple_d_wmma_cshuffle.hpp
Merging the gfx12 code into public repo. (#1362)
2024-06-27 00:33:34 -07:00
device_gemm_multiple_d_xdl_cshuffle_lds_direct_load.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_multiple_d_xdl_cshuffle_v3_ab_scale.hpp
Add KBatch support for gemm_ab_scale (#2740)
2025-10-09 08:33:16 +02:00
device_gemm_multiple_d_xdl_cshuffle_v3_b_preshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_multiple_d_xdl_cshuffle_v3_blockscale_bpreshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_multiple_d_xdl_cshuffle_v3.hpp
add the check of granularity for atomic add (#2959)
2025-10-02 11:15:24 -07:00
device_gemm_multiple_d_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_reduce_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_wmma_cshuffle_v3_b_scale.hpp
Wmma support for multiple ABD GEMM (#2803)
2025-09-22 18:49:06 -07:00
device_gemm_wmma_cshuffle_v3_common.hpp
Wmma support for multiple ABD GEMM (#2803)
2025-09-22 18:49:06 -07:00
device_gemm_wmma_cshuffle_v3.hpp
Wmma support for multiple ABD GEMM (#2803)
2025-09-22 18:49:06 -07:00
device_gemm_wmma_cshuffle_v3r1.hpp
fix compilation errors on RHEL8 and SLES15 (#2967)
2025-10-03 07:08:49 -07:00
device_gemm_wmma.hpp
Merging the gfx12 code into public repo. (#1362)
2024-06-27 00:33:34 -07:00
device_gemm_xdl_cshuffle_lds_direct_load.hpp
TF32 POC in Conv3d on MI30x platform #2763 (second attempt) (#2852)
2025-09-17 14:50:15 -07:00
device_gemm_xdl_cshuffle_streamk_v3.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_xdl_cshuffle_v2.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_xdl_cshuffle_v3_b_preshuffle.hpp
Fix splitk preshuffle (#3137)
2025-11-03 11:59:01 -08:00
device_gemm_xdl_cshuffle_v3_b_scale.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_xdl_cshuffle_v3_mx.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_xdl_cshuffle_v3.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_xdl_cshuffle_v3r1.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_xdl_layernorm_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_xdl_skip_b_lds.hpp
[CK] Fix misc issues in CK examples (#2890)
2025-09-24 11:28:20 -07:00
device_gemm_xdl_splitk_c_shuffle_lds_direct_load.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_xdl_splitk_c_shuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_xdl_streamk.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_xdl_waveletmodel_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_gemm_xdl.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_grouped_contraction_multiple_d_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_grouped_conv_bwd_data_multiple_d_wmma_cshuffle.hpp
Move SetZero functions inside the kernels for Grouped Conv (#2255)
2025-06-11 23:41:03 +02:00
device_grouped_conv_bwd_data_multiple_d_xdl_cshuffle_v1.hpp
Fixed handling of split-K autodeduce argument for grouped convolution (#3024)
2025-10-17 15:36:39 +03:00
device_grouped_conv_bwd_weight_dl.hpp
Remove !defined(__HIP_DEVICE_COMPILE__) in CK kernel (#2564)
2025-07-28 13:01:07 -07:00
device_grouped_conv_bwd_weight_explicit_xdl.hpp
Disable bwd weight split-k autodeduce for single stage kernels (#2856)
2025-09-19 16:27:50 +02:00
device_grouped_conv_bwd_weight_multiple_d_xdl_cshuffle.hpp
Conv:TF32: add more instances - 2 (#2879)
2025-10-10 15:28:17 +08:00
device_grouped_conv_bwd_weight_two_stage_xdl_cshuffle.hpp
Add 2GB limitation for grouped conv bwd weight (#3054)
2025-11-01 14:16:45 +01:00
device_grouped_conv_bwd_weight_wmma_cshuffle.hpp
Fix grid size calc for bwd wei (#2226)
2025-05-26 16:51:09 +02:00
device_grouped_conv_bwd_weight_xdl_cshuffle_v3.hpp
Add 2GB limitation for grouped conv bwd weight (#3054)
2025-11-01 14:16:45 +01:00
device_grouped_conv_bwd_weight_xdl_cshuffle.hpp
Conv:TF32: add more instances - 2 (#2879)
2025-10-10 15:28:17 +08:00
device_grouped_conv_fwd_dl_multiple_d_nhwc_kyxc_nhwk.hpp
Add the last two forward instance traits. (#3134)
2025-10-31 07:52:42 -07:00
device_grouped_conv_fwd_dl_nhwc_kyxc_nhwk.hpp
Remove !defined(__HIP_DEVICE_COMPILE__) in CK kernel (#2564)
2025-07-28 13:01:07 -07:00
device_grouped_conv_fwd_multiple_abd_xdl_cshuffle_v3.hpp
Grouped conv fwd with direct load (#3082)
2025-10-29 09:54:42 +01:00
device_grouped_conv_fwd_multiple_abd_xdl_cshuffle.hpp
Add instance traits for two more grouped forward convolutions (#3112)
2025-10-29 16:04:13 +01:00
device_grouped_conv_fwd_multiple_d_multiple_r_xdl_cshuffle.hpp
[CK] Fix misc issues in CK examples (#2890)
2025-09-24 11:28:20 -07:00
device_grouped_conv_fwd_multiple_d_multiple_r.hpp
update copyright headers (#726)
2023-05-31 18:46:57 -05:00
device_grouped_conv_fwd_multiple_d_wmma_cshuffle.hpp
Add the last two forward instance traits. (#3134)
2025-10-31 07:52:42 -07:00
device_grouped_conv_fwd_multiple_d_xdl_cshuffle.hpp
Add instances for conv_scale with fp8@bf8->fp8 (#1220)
2024-04-03 09:08:08 -05:00
device_grouped_conv_fwd_multiple_d_xdl_large_tensor_cshuffle.hpp
Add instance traits for two more grouped forward convolutions (#3112)
2025-10-29 16:04:13 +01:00
device_grouped_conv_utils.hpp
Add support for GKCYX grouped conv fwd (#2015)
2025-03-26 21:13:38 +01:00
device_grouped_gemm_multi_abd_xdl_fixed_nk.hpp
Fixed handling of split-K autodeduce argument for grouped convolution (#3024)
2025-10-17 15:36:39 +03:00
device_grouped_gemm_multiple_d_dl.hpp
Remove !defined(__HIP_DEVICE_COMPILE__) in CK kernel (#2564)
2025-07-28 13:01:07 -07:00
device_grouped_gemm_multiple_d_splitk_xdl_cshuffle_two_stage.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_grouped_gemm_multiple_d_xdl_cshuffle_tile_loop.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_grouped_gemm_softmax_gemm_permute_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_grouped_gemm_xdl_fixed_nk.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_grouped_gemm_xdl_splitk_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_grouped_gemm_xdl.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_grouped_query_attention_forward_wmma.hpp
Remove !defined(__HIP_DEVICE_COMPILE__) in CK kernel (#2564)
2025-07-28 13:01:07 -07:00
device_image_to_column_impl.hpp
Codegen hipRTC compilation (#1579)
2025-01-31 09:48:39 -08:00
device_max_pool_bwd_impl.hpp
Refactor elementwise kernels (#1222)
2024-04-19 13:31:17 +02:00
device_moe_gemm_blockscale.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_moe_gemm.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_moe_mx_gemm_bns.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_moe_mx_gemm_bpreshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_moe_mx_gemm.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
device_multi_query_attention_forward_wmma.hpp
Remove !defined(__HIP_DEVICE_COMPILE__) in CK kernel (#2564)
2025-07-28 13:01:07 -07:00
device_multiple_reduce_multiblock.hpp
update copyright headers (#726)
2023-05-31 18:46:57 -05:00
device_multiple_reduce_threadwise.hpp
update copyright headers (#726)
2023-05-31 18:46:57 -05:00
device_normalization_bwd_data_impl.hpp
layernorm and groupnorm backward data (#1083)
2023-12-19 04:23:11 +08:00
device_normalization_bwd_gamma_beta_impl.hpp
layernorm and groupnorm backward data (#1083)
2023-12-19 04:23:11 +08:00
device_normalization_fwd_impl.hpp
layernorm and groupnorm backward data (#1083)
2023-12-19 04:23:11 +08:00
device_normalization_fwd_splitk_impl.hpp
layernorm and groupnorm backward data (#1083)
2023-12-19 04:23:11 +08:00
device_permute_impl.hpp
update copyright headers (#726)
2023-05-31 18:46:57 -05:00
device_pool2d_fwd_nhwc_nhwc.hpp
Rewrite pool2d fwd (#1462)
2024-09-11 15:21:00 +02:00
device_pool3d_fwd_ndhwc_ndhwc.hpp
Refactor pool fwd (#815)
2023-08-15 02:25:28 +08:00
device_put_element_impl.hpp
Maxpool bwd (#750)
2023-06-19 09:44:22 -05:00
device_reduce_common.hpp
Support large: 12d tensor size for reduction kenrel (#1465)
2024-08-13 16:15:47 +02:00
device_reduce_multiblock.hpp
Support large: 12d tensor size for reduction kenrel (#1465)
2024-08-13 16:15:47 +02:00
device_reduce_threadwise_multi_d.hpp
Support large: 12d tensor size for reduction kenrel (#1465)
2024-08-13 16:15:47 +02:00
device_reduce_threadwise.hpp
Support large: 12d tensor size for reduction kenrel (#1465)
2024-08-13 16:15:47 +02:00
device_softmax_impl.hpp
Revert "Grouped Gemm with looping over the tiles. (#788)" (#982)
2023-10-11 14:27:29 -05:00
device_sparse_embeddings_forward_layernorm.hpp
update the switch condition for buffer built-ins (#2602)
2025-08-01 14:30:07 -07:00
device_splitk_contraction_multiple_d_xdl_cshuffle.hpp
Extend XDL kernel to Support RDNA3/4 - Part 3 (#2723)
2025-09-09 11:22:36 +08:00
split_k_arg.hpp
Automatic deduction of split-K value for grouped convolution (#2491)
2025-07-31 12:08:45 +02:00
split_k_utils.hpp
Automatic deduction of split-K value for grouped convolution (#2491)
2025-07-31 12:08:45 +02:00