b10018

ggml-org/llama.cppb10018Jul 15, 2026by github-actions[bot]same day as b10017

What changed since b10017

low risk

b10018 adds a SYCL implementation of the xielu activation op; otherwise identical to b10017.

New features

  • SYCL backend now supports the xielu operation, enabling that kernel to run on Intel GPUs instead of falling back.

Upgrade notes

This is a single-commit build bump with no config or API changes, so a straight swap of binaries is sufficient. Only SYCL users relying on xielu-accelerated models will see behavioral differences; other backends are unaffected. Note that the SYCL USM buffer threshold increase (1 GiB to 4 GiB) landed in b10017, not this release.

Analyzed Oct 31, 2026

Upgrading?

Release notes summary

Implements a new `xielu` operation for the SYCL backend, expanding the library's capability for specific computational kernels.

Key Highlights

  • SYCL implementation of the xielu operation
  • Standard release binaries for all supported platforms

Release Notes (excerpt)

<details open> sycl : implement xielu op (#25550) </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:** - Windows x64 (CPU)…

Published by ggml-org on GitHub.

Read the full release notes

Related Repositories