EngineX-Ascend/enginex-ascend-910-llama.cpp
10
0
Fork 0
Code Issues Pull Requests Actions 4 Projects Releases Wiki Activity

musa: enable fp16 mma (all) and cublas on qy2 (#13842)

Browse Source 
* musa: enable fp16 mma (all) and cublas on qy2

Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>

* Update ggml/src/ggml-cuda/ggml-cuda.cu

Co-authored-by: Johannes Gäßler <johannesg@5d6.de>

* Address review comments

Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>

* Address review comments

Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>

* musa: disable MUL_MAT_ID (q2_k × f32) due to precision issues

Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>

---------

Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>
Co-authored-by: Johannes Gäßler <johannesg@5d6.de>
This commit is contained in:
R0CKSTAR
2025-06-26 12:11:59 +08:00
committed by GitHub
parent 60ef23d6c1
commit 716301d1b0
4 changed files with 34 additions and 24 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
ggml/src/ggml-cuda/fattn-wmma-f16.cu
View File

@@ -9,7 +9,11 @@
#ifdef FP16_MMA_AVAILABLE
#if !(defined(GGML_USE_HIP) && defined(__HIP_PLATFORM_AMD__))
#include <mma.h>
#ifdef GGML_USE_MUSA
namespace wmma = mtmusa::wmma;
#else // GGML_USE_MUSA
namespace wmma = nvcuda::wmma;
#endif // GGML_USE_MUSA
#elif defined(GGML_HIP_ROCWMMA_FATTN) && defined(FP16_MMA_AVAILABLE)
#undef HIP_ENABLE_WARP_SYNC_BUILTINS // conflicts with rocWMMA headers
#include <rocwmma/rocwmma.hpp>