Qwen3.5 397B RL completed in 22.78 minutes on 256 GPUs
Company: CoreWeave
The claim, verbatim
CoreWeave completed Qwen3.5 397B reinforcement learning in 22.78 minutes on 256 GB200 GPUs across multiple racks.
Source (primary)
CoreWeave Reports 2.7x Speedup with Vera Rubin Over GB200 on MLPerf DeepSeek-V3 Benchmark - HPCwire (-, news_article)
View cached copy (2026-11-20)Live source ↗Archive.org ↗
Quote: “completed Qwen3.5 397B reinforcement learning in 22.78 minutes on 256 GPUs across multiple racks”
How we checked this
Checked on November 21, 2026. The cited source supports every part of this claim.
The release lists this result under its GB200 NVL72 heading: 22.78 minutes on 256 GPUs across multiple racks.
Confirmed in the source:
- CoreWeave completed Qwen3.5 397B reinforcement learning in 22.78 minutes
- The run used 256 GPUs across multiple racks
- The GPUs were GB200 NVL72
What we did: Read our cached copy of the publisher (https://www.hpcwire.com/aiwire/2026/11/17/coreweave-reports-2-7x-speedup-with-vera-rubin-over-gb200-on-mlperf-deepseek-v3-benchmark/) in full (13,089 characters, retrieved November 20, 2026) and checked each assertion in the claim against it.
Additional evidence
confirms CoreWeave Reports 2.7x Speedup with Vera Rubin Over GB200 on MLPerf DeepSeek-V3 Benchmark - HPCwire
Quote: “completed Qwen3.5 397B reinforcement learning in 22.78 minutes on 256 GPUs across multiple racks”