CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-06 09:46:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

CUDA/HIP: add support for selectable warp size to mmv (#11519)

Browse Source 
CUDA/HIP: add support for selectable warp size to mmv
This commit is contained in:
uvos
2025-02-02 22:40:09 +01:00
committed by GitHub
parent 4d0598e144
commit 396856b400
3 changed files with 34 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
ggml/src/ggml-cuda/common.cuh
View File

@@ -176,6 +176,14 @@ static constexpr bool new_mma_available(const int cc) {
return cc < GGML_CUDA_CC_OFFSET_AMD && cc >= GGML_CUDA_CC_TURING;
}
static constexpr __device__ int ggml_cuda_get_physical_warp_size() {
#if defined(GGML_USE_HIP) && defined(__HIP_PLATFORM_AMD__)
return __AMDGCN_WAVEFRONT_SIZE;
#else
return 32;
#endif // defined(GGML_USE_HIP) && defined(__HIP_PLATFORM_AMD__)
}
[[noreturn]]
static __device__ void no_device_code(
const char * file_name, const int line, const char * function_name, const int arch, const char * arch_list) {