Vera Rubin 2.7x faster than GB200 NVL72 on DeepSeek-V3
Company: CoreWeave
The claim, verbatim
At the same 256-GPU count, CoreWeave's Vera Rubin result trained DeepSeek-V3 2.7x faster than its own GB200 NVL72 submission (CoreWeave-derived, not verified by MLCommons).
Source (primary)
CoreWeave Becomes First Cloud Provider to Deliver Multi-Rack NVIDIA Vera Rubin Results in MLPerf® Training v6.1 - CoreWeave (-, news_article)
View cached copy (2026-11-21)Live source ↗Archive.org ↗
Quote: “At the same 256-GPU count, CoreWeave’s Vera Rubin result trained DeepSeek-V3 2.7x faster than its own GB200 NVL72 submission”
How we checked this
This claim has not yet been checked assertion-by-assertion against its source. It carries a cited source and quote, but the deeper check has not run. When it does, the result appears here whatever it says.
Additional evidence
confirms CoreWeave Becomes First Cloud Provider to Deliver Multi-Rack NVIDIA Vera Rubin Results in MLPerf® Training v6.1 - CoreWeave
Quote: “At the same 256-GPU count, CoreWeave’s Vera Rubin result trained DeepSeek-V3 2.7x faster than its own GB200 NVL72 submission”