WMMA support for GEMM reduce (#2823)

Added gemm + reduce instance library for RDNA4. This includes:

- New device implementation running GEMM and reduction kernel
- instances for wmma (xdl parity)
- examples for wmma (xdl parity)
- tests for existing xdl and wmma

[ROCm/composable_kernel commit: b25d4d684a]
This commit is contained in:
Wojciech Laskowski
2025-09-12 21:36:43 +02:00
committed by GitHub
parent 8c0cdebe63
commit f2edb06bb0
27 changed files with 1911 additions and 89 deletions

View File

@@ -27,3 +27,16 @@ add_example_executable(example_gemm_xdl_splitk_reduce_multi_d_bf16 gemm_xdl_spli
add_example_executable(example_gemm_xdl_splitk_reduce_bf16A_i8B gemm_xdl_splitk_reduce_bf16A_i8B.cpp)
add_example_executable(example_gemm_xdl_splitk_reduce_bfp16 gemm_xdl_splitk_reduce_bf16.cpp)
add_custom_target(example_splitK_gemm_wmma)
add_example_executable(example_gemm_wmma_splitk_reduce_bf16 gemm_wmma_splitk_reduce_bf16.cpp)
add_example_dependencies(example_splitK_gemm_wmma example_gemm_wmma_splitk_reduce_bf16)
add_example_executable(example_gemm_wmma_splitk_reduce_bf16A_i8B gemm_wmma_splitk_reduce_bf16A_i8B.cpp)
add_example_dependencies(example_splitK_gemm_wmma example_gemm_wmma_splitk_reduce_bf16A_i8B)
add_example_executable(example_gemm_wmma_splitk_reduce_multi_d_bf16 gemm_wmma_splitk_reduce_multi_d_bf16.cpp)
add_example_dependencies(example_splitK_gemm_wmma example_gemm_wmma_splitk_reduce_multi_d_bf16)
add_example_executable(example_gemm_wmma_splitk_reduce_multi_d_fp16 gemm_wmma_splitk_reduce_multi_d_fp16.cpp)
add_example_dependencies(example_splitK_gemm_wmma example_gemm_wmma_splitk_reduce_multi_d_fp16)