ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-07-13 02:27:33 +00:00
Code Issues Packages Projects Releases Wiki Activity

temp work saved, changed the BDataType to f16 or bf16 since wmma currently not support non-equal A and B datatype

Browse Source 
(cherry picked from commit 22fbd68f1db458ab50780a394ee2544c7a1484d1)

Co-authored-by: Cenxuan <cenxuan@streamhpc.com>
This commit is contained in:
Apoorva Kalyani
2025-05-26 21:36:36 +00:00
committed by apoorva
parent b129e731c3
commit 455275de80
5 changed files with 46 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
include/ck/tensor_operation/gpu/warp/wmma_gemm.hpp
View File

@@ -16,11 +16,13 @@ enum struct WmmaInstr
wmma_f32_16x16x16_bf16,
wmma_f16_16x16x16_f16,
wmma_bf16_16x16x16_bf16,
wmma_i32_16x16x16_iu16,
wmma_i32_16x16x16_iu8,
wmma_i32_16x16x16_iu4,
// gfx12
wmma_f32_16x16x16_f16_gfx12,
wmma_f32_16x16x16_bf16_gfx12,
wmma_i32_16x16x16_iu16_gfx12,
wmma_i32_16x16x16_iu8_gfx12,
wmma_f32_16x16x16_f8f8_gfx12,
wmma_f32_16x16x16_f8bf8_gfx12,
@@ -590,6 +592,16 @@ struct WmmaSelector
return WmmaInstr::wmma_bf16_16x16x16_bf16;
}
template <>
constexpr auto GetWmma<unsigned short, unsigned short, int, 16, 16>()
{
#ifdef __gfx12__
return WmmaInstr::wmma_i32_16x16x16_iu16_gfx12;
#else
return WmmaInstr::wmma_i32_16x16x16_iu16;
#endif
}
template <>
constexpr auto GetWmma<int8_t, int8_t, int, 16, 16>()
{