CoreWeave powers up multi-rack NVIDIA Vera Rubin cluster as shares slip

CoreWeave said it has brought a multi-rack cluster based on NVIDIA’s Vera Rubin architecture online, marking one of the first operational deployments of the next-generation AI computing platform. The announcement comes as the cloud infrastructure provider’s stock trades down 3.09% on the day, changing hands at $80.92 after closing the prior session at $83.50.

The Livingston, New Jersey-based company, which specializes in AI-focused cloud services, said the deployment underscores its ability to stand up large-scale computing environments quickly as demand for training and inference capacity continues to grow. Multi-rack clusters of this type require substantial power, cooling, and networking coordination, and rapid deployment timelines have become a competitive differentiator among AI cloud providers.

CoreWeave’s CoreWeave Cloud platform combines proprietary software with cloud services designed to automate the management of complex AI workloads. The company positions itself as an alternative to hyperscale cloud providers for enterprises and AI labs seeking specialized GPU capacity.

The market cap stands at roughly $48.36 billion, and shares have faced pressure in recent sessions despite a steady stream of announcements around new data center capacity and hardware deployments. The stock’s pullback to $80.92 on Thursday came even as the company highlighted the Vera Rubin milestone, which is being closely watched as an early indicator of how quickly the industry can absorb NVIDIA’s latest architecture.

Vera Rubin represents NVIDIA’s next step beyond the Blackwell generation, and early deployments by AI cloud specialists like CoreWeave are often viewed as bellwethers for the broader AI infrastructure build-out cycle. The company did not disclose the exact scale of the cluster or customer commitments tied to it in the release.

What to watch

• CoreWeave’s next quarterly earnings report and any updates on capacity expansion and revenue tied to next-generation GPU deployments
• Further details on the size of the Vera Rubin cluster and which customers will utilize it
• Broader rollout pace of NVIDIA’s Rubin architecture across cloud providers

Source: original release