CoreWeave (CRWV) shares advanced on Wednesday after the AI cloud provider announced the deployment of a multi-rack Nvidia Vera Rubin NVL72 cluster on its CoreWeave Cloud platform. The setup integrates hundreds of Nvidia Rubin GPUs into a single scale-out system, enabling customers to run both training and inference workloads across multiple racks.
The new infrastructure is designed to support larger AI models, more demanding inference tasks, and reinforcement learning at scale. CoreWeave stated that this configuration provides additional capacity for customers developing agentic AI, allowing for greater scale and faster iteration.
According to the company, a single Nvidia Vera Rubin NVL72 rack combines 72 Rubin GPUs with 36 Vera CPUs, along with Nvidia networking and data-processing technology. Multiple racks can be interconnected via Nvidia Spectrum-X Ethernet to function as one cohesive cluster. CoreWeave claims to be the first AI cloud provider to validate and bring up a Vera Rubin NVL72 system.
Chen Goldberg, Executive Vice President of Product and Engineering at CoreWeave, highlighted that the multi-rack configuration connects hundreds of Rubin GPUs as a single scale-out cluster. The company also noted that its infrastructure automates rack setup, firmware updates, validation, power, and cooling processes, with full rack testing conducted before production deployment.
Despite the positive news, CoreWeave's stock pared its gains after the Federal Reserve raised its benchmark interest rate to 4% in a unanimous decision, signaling further hikes. The stock had risen as much as 4% before settling around 3% higher.
In a separate development, CoreWeave also expanded its AI Object Storage platform with several new features. These include cross-region write acceleration, which allows AI workloads to write data in the region where they are running while CoreWeave copies the data to another region in the background, reducing pauses for multi-region training. The company also introduced an Archive tier with no retrieval, early-deletion, or read fees, and a Local Object Transport Accelerator (LOTA) that caches data closer to AI workloads, cutting read latency by eight times compared to traditional storage clusters.
CoreWeave also received a boost from a commercial agreement with Blockfusion USA's subsidiary North East Data, which converted a non-binding letter of intent into a definitive 15-year anchor lease for capacity at its Niagara Falls, New York campus. The agreement includes two five-year renewal options and a companion expansion agreement.
However, analysts at Bernstein expressed caution in a September 14 report, noting that CoreWeave could be among the companies most exposed if AI firms reduce spending on model training. The firm highlighted CoreWeave's heavy reliance on AI infrastructure and its concentration in rural and lower-tier data-center markets as potential risks. Bernstein maintained a Sell rating on CoreWeave while raising its price target to $74.
Investors may also be watching broader AI infrastructure trends, as neocloud stocks have slid despite surging revenue, and Nvidia shares have dipped on AI slowdown fears. The Fed's rate hike also impacts growth-oriented tech names, as seen in KOSPI's rebound on chip gains.
This article is for informational purposes only and does not constitute financial advice.
