Skip to content

[XPU] fix float32 matmul precision: use FC_FLOAT default instead of FC_TF32 on XRE5 #78625

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

Open
YqGe585 wants to merge 1 commit into
base: develop
Choose a base branch
from
+4 −1
Open

[XPU] fix float32 matmul precision: use FC_FLOAT default instead of FC_TF32 on XRE5 #78625

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
5 changes: 4 additions & 1 deletion paddle/phi/kernels/xpu/xpu_api_wrapper.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -64,7 +64,10 @@ inline XPUFCCalcType FCCalcType() {
{"XPU_PADDLE_FC_INT32_WITH_LL", XPUFCCalcType::FC_INT32_WITH_LL},
};
#ifdef PADDLE_WITH_XPU_XRE5
auto default_calc_type = XPUFCCalcType::FC_TF32;
// Use full float32 accumulation by default for better precision, matching
// the GPU default (FLAGS_cublas_allow_tf32=false). Users who need TF32
// performance can set env var XPU_PADDLE_FC_TF32.
auto default_calc_type = XPUFCCalcType::FC_FLOAT;
#else
auto default_calc_type = XPUFCCalcType::FC_INT16;
#endif
Expand Down
Loading