The llama.cpp project has released version b11752, introducing improvements to its multi-expert (MoE) routing selection and device identification on Apple Silicon. The update refines MoE router scoring to ensure consistent expert selection, resolving issues observed on Apple M2 Ultra chips. Additionally, it enhances the device table to differentiate between CPU and Metal GPU on Apple SoCs, improving test clarity. The release also includes support for various platforms like macOS, Linux, Windows, and openEuler. AI
IMPACT Improves performance and testing for local LLM inference on Apple hardware.
RANK_REASON This is a software release for a specific project (llama.cpp) with technical improvements, not a frontier model release or significant industry event.
Read on llama.cpp — Releases →
- ACL Graph
- Apple M2 Ultra
- Apple Silicon
- .claude
- KleidiAI
- llama.cpp
- M5 Max
- MacBook
- Metal
- OpenEuler
- Qwen3.8-Flash-Next-GGUF
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →