AI tools & infrainfo
Kimi K3 (2.8T) at 1 token/s on a MacBook Pro, streamed from four SSDs
AI summary · glm-5.3-flashSummaries by Vani AI
A GitHub project streams Moonshot's 2.8T-parameter Kimi K3 from four SSDs to run at 1 token/s on a MacBook Pro.
A GitHub repository (argonautlabsai/deltafin) demonstrates running Moonshot AI's Kimi K3, a 2.8-trillion-parameter model, on a MacBook Pro by streaming its weights from four SSDs at 1 token per second. The project drew 55 points and 22 comments on Hacker News, highlighting consumer-hardware inference for extremely large models.
Key facts
- Kimi K3 has 2.8 trillion parameters
- Inference runs at 1 token/s on a MacBook Pro
- Weights are streamed from four SSDs
- Project is hosted at GitHub repo argonautlabsai/deltafin
- Vendors
- Moonshot AI
- Products
- MacBook Pro
- AI models
- Kimi K3
From the original article
By Argonautlabs · github.com
This source shares headlines only. Summary, key facts and assessment above are AI-written from the article and can be wrong.