llama.cpp PR runs missed MoE experts on CPU

llama.cpp pull request #30966 adds llama_moe_cache, which runs missed MoE experts on the CPU, plus a tuning script. It is seen as a significant speedup for low-VRAM users offloading MoE models.

• llama.cpp Adds Tensor Merge on GGUF Conversion2026-10-26
• Community tests full n-gram table for Qwen 3.8 Flash Next2026-10-18
• llama.cpp Adds GPU Cache for MoE Experts in Host Memory2026-10-07
• User Shares Strata Speedup for Local LLMs but 64GB RAM Constraint2026-10-04
• 176B MoE Model Runs on 16GB Laptop GPU2026-10-04