Fastest 72-GPU Llama 3.1 8B submission at 7.16 minutes
Company: CoreWeave
The claim, verbatim
On a single Vera Rubin NVL72 rack, CoreWeave reached target quality on Llama 3.1 8B in 7.16 minutes, the fastest 72-GPU submission, 1.9x faster than its GB200 NVL72 submission (13.70 minutes).
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: “On a single NVIDIA Vera Rubin NVL72 rack, CoreWeave reached target quality in 7.16 minutes, the fastest 72-GPU submission. That is 1.9x faster than its 72-GPU GB200 NVL72 submission (13.70 minutes).”
How we checked this
Checked on November 21, 2026. The cited source supports every part of this claim.
The release names Llama 3.1 8B and gives the 7.16-minute result, the fastest-72-GPU status, and the 1.9x comparison with 13.70 minutes.
Confirmed in the source:
- On a single Vera Rubin NVL72 rack, CoreWeave reached target quality on Llama 3.1 8B in 7.16 minutes
- This was the fastest 72-GPU submission
- This was 1.9x faster than CoreWeave's GB200 NVL72 submission of 13.70 minutes
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: “On a single NVIDIA Vera Rubin NVL72 rack, CoreWeave reached target quality in 7.16 minutes, the fastest 72-GPU submission. That is 1.9x faster than its 72-GPU GB200 NVL72 submission (13.70 minutes).”