Skip to content

perf(qwen-flash): fence each K block in the W8A8 WMMA GEMM #255

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
fedeizzo merged 2 commits into from
Sep 25, 2026
Merged

perf(qwen-flash): fence each K block in the W8A8 WMMA GEMM #255

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
554199a
perf(qwen-flash): fence each K block in the W8A8 WMMA GEMM
pwilkin Sep 24, 2026
893fb98
perf(qwen-flash): fence W8A8 K blocks only on LLVM 23+
pwilkin Sep 24, 2026
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 src/models/qwen38_flash_next/kernels/rocm/kernels.hip.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3530,6 +3530,12 @@ __launch_bounds__(256) __global__ void W8A8BlockedWmmaGEMMKernel(
}
}
}
#if __clang_major__ >= 23
// LLVM 23 hoists the next K block's operands above this one's WMMAs,
// which spills registers on gfx1151. ROCm 7.2.3 (LLVM 22) does not, and
// the fence costs it about 3%.
__builtin_amdgcn_sched_barrier(0);
#endif
}
__syncthreads();
}
Expand Down