Skip to content

Setting GEMM_PREFERED_SIZE parameter for A64FX #5348

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
martin-frbg merged 1 commit into from
Jun 27, 2025
+6 −0
Merged

Setting GEMM_PREFERED_SIZE parameter for A64FX #5348

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
6 changes: 6 additions & 0 deletions param.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3701,6 +3701,12 @@ is a big desktop or server with abundant cache rather than a phone or embedded d

#elif defined(A64FX) // 512-bit SVE

#if defined(XDOUBLE) || defined(DOUBLE)
#define GEMM_PREFERED_SIZE 8
#else
#define GEMM_PREFERED_SIZE 16
#endif

/* When all BLAS3 routines are implemeted with SVE, SGEMM_DEFAULT_UNROLL_M should be "sve_vl".
Until then, just keep it different than DGEMM_DEFAULT_UNROLL_N to keep copy routines in both directions seperated. */
#define SGEMM_DEFAULT_UNROLL_M 4
Expand Down
Loading