CoreWeave has become the first company to power up Nvidia's Vera Rubin NVL72 rack-scale system, completing operational validation on June 1, 2026, a day after Dell Technologies delivered the hardware. The new rack delivers up to ten times the token throughput per megawatt of Nvidia's previous Blackwell platform, and Nvidia expects the system to reach full production by July 2026.
CoreWeave completed operational validation of the Vera Rubin NVL72 on June 1, 2026, one day after Dell Technologies delivered what amounts to the most powerful commercially deployed AI inference machine ever built. Each rack packs 72 Rubin GPUs and 36 Vera CPUs. It is linked by NVLink 6 fabric running at 260 TB/s and is capable of 3.6 exaFLOPS of NVFP4 inference capability.
A tenfold jump in throughput
The Vera Rubin NVL72 delivers up to 10x better token throughput per megawatt on complex workloads like DeepSeek R1 compared with Blackwell. Fewer GPUs are therefore needed to handle equivalent workloads, which lowers the cost per million tokens for large-scale inference. Dell also moved fast on integration: the team took the system from physical delivery to production readiness in under 6.5 hours.
CoreWeave's market reaction
CoreWeave and Dell have built on a partnership spanning prior NVL system generations, and CoreWeave's stock spiked roughly 14% following the announcement.
What comes next
Nvidia expects the Vera Rubin platform to enter full production by July 2026. Deployments are planned across OpenAI, Google Cloud, and Microsoft Azure.
Source: Crypto Briefing
Trading involves risk.
Risk Disclaimer
