CoreWeave Becomes First Cloud Provider to Deliver Multi-Rack NVIDIA Vera Rubin Results in MLPerf® Training v6.1
Business Wire•16/11/2026•12:00 ET
17
Key Highlights
- ➤17.65-minute DeepSeek-V3 671B training on 256 NVIDIA Rubin GPUs across multiple racks
- ➤CoreWeave was the only cloud provider submitting multi-rack NVIDIA Rubin results
- ➤Vera Rubin trained DeepSeek-V3 2.7x faster than CoreWeave’s GB200 submission
- ➤7.16-minute Llama 3.1 8B result was fastest 72-GPU submission
- ➤GB200 trained GPT-OSS 20B in 17.18 minutes on one 72-GPU rack
Expert Statements
Chen Goldberg, Executive vice president of product and engineering at CoreWeave
“New silicon is only as fast as the cloud it runs on”
Chen Goldberg, Executive vice president of product and engineering at CoreWeave
“Our MLPerf Training v6.1 results show what NVIDIA Vera Rubin can do when the full stack is built for it, from a single rack to multi-rack scale.”
Chen Goldberg, Executive vice president of product and engineering at CoreWeave
“We ran these benchmarks on the same production clusters our customers use to train frontier models, with CoreWeave Mission Control keeping the fleet healthy throughout.”
CoreWeave Becomes First Cloud Provider to Deliver Multi-Rack NVIDIA Vera Rubin Results in MLPerf® Training v6.1
The only cloud provider to submit multi-rack NVIDIA Vera Rubin NVL72 results, CoreWeave trained DeepSeek-V3 671B 2.7x faster than its submission on NVIDIA GB200 NVL72 at the same GPU count
Get started
Create a free account to read the full story.
