happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

faster mmf

This commit is contained in:
zhang hui 2026-01-17 21:08:16 +08:00
parent 7444a7a18b
commit 250ae9aee8
1 changed files with 0 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
ggml/src/ggml-cuda/mma.cuh
View File

@ -993,17 +993,10 @@ namespace ggml_cuda_mma {
using floatx4_t = __attribute__((ext_vector_type(4))) float;
floatx4_t& acc_frag = reinterpret_cast<floatx4_t&>(D.x[0]);
#if defined(CDNA3)
#if 0
using floatx2_t = __attribute__((ext_vector_type(2))) float;
const floatx2_t& a_frag = reinterpret_cast<const floatx2_t&>(A.x[0]);
const floatx2_t& b_frag = reinterpret_cast<const floatx2_t&>(B.x[0]);
acc_frag = __builtin_amdgcn_mfma_f32_16x16x8_xf32(a_frag, b_frag, acc_frag, 0, 0, 0);
#else
#pragma unroll
for (int i = 0; i < 2; ++i) {
acc_frag = __builtin_amdgcn_mfma_f32_16x16x4f32(A.x[i], B.x[i], acc_frag, 0, 0, 0);
}
#endif
#elif defined(CDNA2) || defined(CDNA1)
#pragma unroll
for (int i = 0; i < 2; ++i) {