The Bajoran Engineer

Posts tagged apple-silicon Clear

1. Oct 17, 2026 · 3 min read

   TIL: MLX is only a little faster than llama.cpp on my M1

   This is part one, on an M1 with 16 GB of memory. In part two I'll run the exact same test on my 10-core M5 with 24 GB.

   • #til
   • #mlx
   • #llama-cpp
   • #apple-silicon
   • #local-llm
   • #benchmarks
   • #AI
   • #LLM
   • #python
   • #AI-assisted-post