ver4a/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

CUDA: fix FA logic for PTX 7.0 and CC >= 7.5 (#12222)

Browse source
This commit is contained in:
Johannes Gäßler 2025-03-06 18:45:09 +01:00 • committed by GitHub
parent 3ffbbd5ce1
commit 5220a16d18
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ggml/src/ggml-cuda/fattn.cu
View file

@ -310,7 +310,7 @@ void ggml_cuda_flash_attn_ext(ggml_backend_cuda_context & ctx, ggml_tensor * dst
}
// The MMA implementation needs Turing or newer, use the old WMMA code for Volta:
if (cc == GGML_CUDA_CC_VOLTA) {
if (fp16_mma_available(cc) && !new_mma_available(cc)) {
ggml_cuda_flash_attn_ext_wmma_f16(ctx, dst);
return;
}