Skip to content

Add ds_swizzle+ds_bpermute cross-lane reduction for CDNA wave64

cb4dba3
Select commit
Loading
Failed to load commit list.
Open

[DRAFT] Add register-only cross-lane reduction for attention #2359

Add ds_swizzle+ds_bpermute cross-lane reduction for CDNA wave64
cb4dba3
Select commit
Loading
Failed to load commit list.