Change8

b9813

📦 llama-cppView on GitHub →
3 features🐛 2 fixes🔧 1 symbols

Summary

This release updates the Vulkan backend by adding the INTEL_XE1 architecture enum and enabling coopmat1 on Intel Xe-LPG Plus devices, alongside renaming an existing enum.

Migration Steps

  1. Rename the architecture enum from INTEL_PRE_XE2 to INTEL_XE1 if referencing it directly in code.

✨ New Features

  • Vulkan backend: Added INTEL_XE1 architecture enum.
  • Vulkan backend: Enabled coopmat1 support on Intel Xe-LPG Plus hardware.
  • Added Windows driver check for Vulkan support.

🐛 Bug Fixes

  • Addressed comments regarding bf16 usage.
  • Removed trailing whitespace.

Affected Symbols