b9886

ggml-org/llama.cppb9886Jul 6, 2026by github-actions[bot]same day as b9885

What changed since b9885

low risk

b9886 adds an ARM CPU optimization using a UE4M3 lookup table for NVFP4 dot products, with no other changes noted.

Improvements

  • ARM CPU backend now uses a UE4M3 lookup table for NVFP4 dot product operations, improving computational efficiency on ARM hardware.

Upgrade notes

This is a single-commit increment over b9885 with no config or API changes; a drop-in binary swap is sufficient. ARM users running NVFP4-quantized models are the primary beneficiaries, while other platforms see no functional difference.

Analyzed Oct 28, 2026

Upgrading?

Release notes summary

The CPU backend has been optimized for ARM architecture by utilizing the UE4M3 Look-Up Table (LUT) for ARM NVFP4 dot product operations, aiming to improve computational efficiency.

Key Highlights

  • Use UE4M3 LUT for ARM NVFP4 dot product
  • Optimize ARM CPU performance

Release Notes (excerpt)

<details open> ggml-cpu: use UE4M3 LUT in ARM NVFP4 dot product (#25331) </details> **macOS/iOS:** - macOS Apple Silicon (arm64) - macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED - macOS Intel (x64) - iOS XCFramework **Linux:** - Ubuntu x64 (CPU) - Ubuntu arm64 (CPU) - Ubuntu s390x (CPU) - Ubuntu x64 (Vulkan) - Ubuntu arm64 (Vulkan) - Ubuntu x64 (ROCm 7.2) - Ubuntu x64 (OpenVINO) - Ubuntu x64 (SYCL FP32) - Ubuntu x64 (SYCL FP16) **Android:** - Android arm64 (CPU) **Windows:**…

Published by ggml-org on GitHub.

Read the full release notes

Related Repositories