Enrico Degregori 0b149c8695 Wmma support for multiple ABD GEMM (#2803) ...

* multi_abd wmma support:

 - Add multiple A and B support to multiple D implementation (gridwise level)
 - Add multi_abd GEMM (device level)
 - Add instances (xdl parity)
 - Add tests (both xdl and wmma)
 - Add examples
 - Add ckProfiler support (both xdl and wmma)

* Fix bug in device print function

* Fix unused template parameter

* Fix batched gemm for multiABD gridwise implementation

* Fix gemm_universal_reduce with multiABDs gridwise implementation

---------

Co-authored-by: Illia Silin <98187287+illsilin@users.noreply.github.com>

[ROCm/composable_kernel commit: 3d29bff2f0]

2025-09-22 18:49:06 -07:00

..

batched_gemm

Implement batched gemm wmma (RDNA batched gemm) based on wmma cshuffle v3 (#2319)

2025-06-24 07:28:13 -07:00

batched_gemm_gemm

Extend XDL kernel to Support RDNA3/4 - Part 4 (#2724)

2025-09-12 08:17:07 -07:00

batched_gemm_multi_d

Refactoring cmake files to build data types separately. (#932)

2023-09-20 22:15:56 -07:00

batched_gemm_reduce

Split the instances by architecture. (#1223)

2024-04-02 09:42:17 -07:00

batched_gemm_softmax_gemm

Extend XDL kernel to Support RDNA3/4 - Part 4 (#2724)

2025-09-12 08:17:07 -07:00

batched_gemm_softmax_gemm_permute

Extend XDL kernel to Support RDNA3/4 - Part 4 (#2724)

2025-09-12 08:17:07 -07:00

batchnorm

Resolve some data type issues and cmake policy. (#940)

2023-09-26 08:39:11 -07:00

block_swizzle_test

Switch to C++20 standard for all CMake targets. (#2536)

2025-07-22 10:52:10 -07:00

block_to_ctile_map

update copyright headers (#726)

2023-05-31 18:46:57 -05:00

ck_tile

[CK_TILE] Tensor-wise scaled quant gemm kernel (#2846)

2025-09-19 16:52:35 -07:00

contraction

Extend XDL kernel to Support RDNA3/4 - Part 4 (#2724)

2025-09-12 08:17:07 -07:00

conv_tensor_rearrange

Fixing most of the cppcheck errors. (#1142)

2024-01-24 13:47:48 -08:00

conv_util

Add Grouped Conv Fwd Large Tensor kernel (#1432)

2024-08-06 10:06:10 +02:00

convnd_bwd_data

Split the instances by architecture. (#1223)

2024-04-02 09:42:17 -07:00

convnd_fwd

Split the instances by architecture. (#1223)

2024-04-02 09:42:17 -07:00

data_type

poc convert fnuz fp8 to non-native dtype similar to ocp (#2871)

2025-09-18 22:51:01 -07:00

elementwise_normalization

Refactoring cmake files to build data types separately. (#932)

2023-09-20 22:15:56 -07:00

gemm

Refactoring cmake files to build data types separately. (#932)

2023-09-20 22:15:56 -07:00

gemm_add

Wmma support for multiple Ds based GEMMs (#2613)

2025-09-05 16:31:08 +02:00

gemm_b_scale

Support b_scale: (#2350)

2025-07-24 18:49:58 -07:00

gemm_layernorm

Extend XDL kernel to Support RDNA3/4 - Part 4 (#2724)

2025-09-12 08:17:07 -07:00

gemm_multi_abd

Wmma support for multiple ABD GEMM (#2803)

2025-09-22 18:49:06 -07:00

gemm_mx

MX GEMM - Add FP6 GEMM Test (#2488)

2025-07-11 15:32:12 -06:00

gemm_reduce

Split the instances by architecture. (#1223)

2024-04-02 09:42:17 -07:00

gemm_split_k

Split the instances by architecture. (#1223)

2024-04-02 09:42:17 -07:00

gemm_universal

Add missing copyright headers (#2359)

2025-06-17 14:29:45 -07:00

gemm_universal_reduce

WMMA support for GEMM reduce (#2823)

2025-09-12 21:36:43 +02:00

gemm_universal_streamk

Add missing copyright headers (#2359)

2025-06-17 14:29:45 -07:00

grouped_convnd_bwd_data

Extend XDL kernel to Support RDNA3/4 - Part 5 (#2725)

2025-09-15 10:59:25 -07:00

grouped_convnd_bwd_weight

Disable bwd weight split-k autodeduce for single stage kernels (#2856)

2025-09-19 16:27:50 +02:00

grouped_convnd_fwd

Extend XDL kernel to Support RDNA3/4 - Part 5 (#2725)

2025-09-15 10:59:25 -07:00

grouped_convnd_fwd_activation

Extend XDL kernel to Support RDNA3/4 - Part 5 (#2725)

2025-09-15 10:59:25 -07:00

grouped_gemm

Extend XDL kernel to Support RDNA3/4 - Part 4 (#2724)

2025-09-12 08:17:07 -07:00

magic_number_division

update copyright headers (#726)

2023-05-31 18:46:57 -05:00

mx_mfma_op

fix test_mx_mfma errors (#2614)

2025-08-04 11:43:47 -07:00

normalization_bwd_data

Grouped GEMM Multiple D tile loop. (#1247)

2024-04-25 15:12:53 -05:00

normalization_bwd_gamma_beta

Grouped GEMM Multiple D tile loop. (#1247)

2024-04-25 15:12:53 -05:00

normalization_fwd

layernorm and groupnorm backward data (#1083)

2023-12-19 04:23:11 +08:00

permute_scale

Split the instances by architecture. (#1223)

2024-04-02 09:42:17 -07:00

pool

upgrade from clang-format-12 to clang-format-18 (#2568)

2025-07-28 11:34:07 -07:00

position_embedding

[CK_TILE] support group from cmdline (#1295)

2024-05-28 11:13:21 +08:00

quantization

Added wmma support for gemm quantization: (#2841)

2025-09-16 16:23:29 -07:00

reduce

Rewrite *sh reduce unit tests to gtest: part 1 (#1407)

2024-08-12 16:28:10 +02:00

reference_conv_fwd

upgrade from clang-format-12 to clang-format-18 (#2568)

2025-07-28 11:34:07 -07:00

scatter_gather

Fix build errors on windows (#2456)

2025-07-16 07:58:23 -07:00

smfmac_op

Merge from internal (#1857)

2025-02-07 15:05:05 -07:00

softmax

[HotFix] add config and version files to pass on build info (#856)

2023-08-23 11:36:17 -07:00

space_filling_curve

update copyright headers (#726)

2023-05-31 18:46:57 -05:00

transpose

Split the instances by architecture. (#1223)

2024-04-02 09:42:17 -07:00

wmma_op

Fix issue with multiple targets and remove smfmac tests from unsupported test targets (#1372)

2024-07-03 23:34:38 -07:00

wrapper

Extend XDL kernel to Support RDNA3/4 - Part 4 (#2724)

2025-09-12 08:17:07 -07:00

CMakeLists.txt

Wmma support for multiple ABD GEMM (#2803)

2025-09-22 18:49:06 -07:00