Single-User Inference Card

(markohaberl.substack.com)

6 points | by mhaberl 13 hours ago ago

2 comments

  • mhaberl 3 minutes ago

    You could run Qwen 3.8 locally at almost 200 tok/s

  • deadcatfound 12 hours ago

    This is the form factor I want: one quiet box for a single operator, not a mini data center. Sustained speed and power draw matter most.