Repository navigation

perf: run large CPU matmuls on Accelerate on Apple Silicon - #29

Merged

danielhanchen merged 22 commits into

perf/mac-cpu-blas

Oct 12, 2026

Merged

perf: run large CPU matmuls on Accelerate on Apple Silicon#29
danielhanchen merged 22 commits into
masterfrom
perf/mac-cpu-blas

Commits on Oct 11, 2026
Commits on Oct 12, 2026
- CPU: hand large matmuls to ggml's BLAS backend when it is built in
  danielhanchen
  committed
  8e47549
- macOS prebuilts: Accelerate BLAS on the arm64 leg, off on the Intel leg
  danielhanchen
  committed
  a38052c
- Trim comments for CPU BLAS scheduling
  danielhanchen
  committed
  8a40827
- CPU BLAS: run eval callbacks on the scheduled CPU path
  danielhanchen
  committed
  d6736ba
- CPU BLAS: release the BLAS conversion scratch at runner end
  danielhanchen
  committed
  0fab806
- Trim comments
  danielhanchen
  committed
  ef6f8c6
- Merge remote-tracking branch 'origin/master' into pr29-head
  danielhanchen
  committed
  ad62867