Skip to content

sycl: pick the src1 fp16 conversion from the runtime oneDNN switch #288

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
bri-prism merged 1 commit into from
Sep 28, 2026
Merged

sycl: pick the src1 fp16 conversion from the runtime oneDNN switch #288

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 5 additions & 2 deletions ggml/src/ggml-sycl/ggml-sycl.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3540,6 +3540,7 @@ static void ggml_sycl_mul_mat_batched_sycl(ggml_backend_sycl_context & ctx, cons
" : converting src1 to fp16");

#if GGML_SYCL_DNNL
if (g_ggml_sycl_enable_dnn) {
// iterate tensor dims and find the slowest moving dim and stride
int last_dim=0;
int last_str=0;
Expand All @@ -3565,13 +3566,15 @@ static void ggml_sycl_mul_mat_batched_sycl(ggml_backend_sycl_context & ctx, cons
const to_fp16_sycl_t to_fp16_sycl = ggml_get_to_fp16_sycl(src1->type, dst);
GGML_ASSERT(to_fp16_sycl != nullptr);
to_fp16_sycl(src1_f16, src1_f16_alloc.get(), ne_src1, queue);
# else
} else
#endif
{
const int64_t ne_src1 = ggml_nelements(src1);
src1_f16_alloc.alloc(ne_src1);
const to_fp16_nc_sycl_t to_fp16_nc_sycl = ggml_get_to_fp16_nc_sycl(src1->type);
GGML_ASSERT(to_fp16_nc_sycl != nullptr);
to_fp16_nc_sycl(src1_f16, src1_f16_alloc.get(), ne10, ne11, ne12, ne13, s11, s12, s13, queue);
#endif
}

src1_f16 = src1_f16_alloc.get();
s11 = ne10;
Expand Down
Loading