Hacker News new | ask | show | jobs
by fragmede 57 days ago
I clocked my M4 at 108 Watts while running inference using Qwen3.6-35b-a3b via Al dente.