Full Kimi K3 model runs on 16x GB10 cluster at 20+ TPS
ciprianveg · reddit · 2026-08-05
A developer has successfully run the full Kimi K3 model on a 16x GB10 cluster. Benchmarks show an average generation speed of over 20 TPS, peaking at 38 TPS, with a prefill speed of 750 TPS.
The author noted this is the first full K3 run using dspark on this cluster. Further optimizations and tests are planned, with the vLLM image and deployment instructions to be released soon.
More from Infra
- Elon Musk Expects First Starmind AI Satellites to Launch Next Year — Polymarket · 2026-08-05
- GPU Rental Costs Surge: Blackwell Clusters Up 60% in Six Months — Alec_Coughlin · 2026-08-05
- Musk: SpaceX to Triple GPU Compute by 2027, Going All-In on Nvidia — firstadopter · 2026-08-05
- Elon Musk says SpaceX committed to using Nvidia GPUs exclusively — elonmusk · 2026-08-05
- 25% Chance of an AI Data Center in Space by 2027, Polymarket Odds Show — Polymarket · 2026-08-05
- Report: SpaceX to Use Nvidia Chips for Orbital AI Data Center Satellites — Polymarket · 2026-08-05