Massive RL Run: ~1,000 GB300 GPUs and $5M?
natolambert · x · 2026-07-16
This reply discussed the potential scale of a post-training RL run, with the author providing a rough estimate: around 1K GB300 GPUs, running for 1 week, and costing roughly $5 million.
They added that while this estimate is an intuitive guess based on Claude—and they would welcome a more rigorous mathematical analysis—Claude's intuition regarding inference costs might not be entirely reliable in agentic settings.
More from Infra
- Vercel AI Gateway data shows Anthropic, OpenAI and Google at 97.09% spend share — cramforce · 2026-07-21
- NVIDIA starts rolling out 102.4 Tbps Spectrum-6 switches for Vera Rubin AI factories — nvidia · 2026-07-21
- Mustafa Suleyman says Microsoft is preparing for an OpenAI exit, while a new chip costs 30% less than GB200 — thoefler · 2026-07-21
- Microsoft and Mistral sign multi-billion-dollar deal to expand AI infrastructure in Europe — The Decoder · 2026-07-21
- Speculative decoding boosts Qwen3.6-27B on one 5090, but slows crowded servers — luke_pacman · 2026-07-21
- NVIDIA says Blackwell Ultra hit 1,648 TFLOPs per GPU on DeepSeek-V3 671B training — NVIDIAAI · 2026-07-21