CoreWeave launches AI GPU cluster

• CoreWeave deployed multi-rack NVIDIA Vera Rubin NVL72 clusters on its AI cloud platform.
• The system connects hundreds of NVIDIA Rubin GPUs into a single AI computing cluster.
• CoreWeave also introduced new AI storage features for cross-region workloads.

CoreWeave (NASDAQ:CRWV) announced the deployment of multi-rack NVIDIA Vera Rubin NVL72 clusters on CoreWeave Cloud, enabling large-scale AI workloads using hundreds of NVIDIA Rubin GPUs in a single cluster.

The company said the infrastructure is designed for agentic AI workloads that require large-scale model training, inference, and repeated data processing across distributed systems.

CoreWeave also introduced new capabilities in its AI Object Storage platform, including cross-region write acceleration and an Archive storage tier designed to reduce delays and storage costs for AI workloads.

The NVIDIA Vera Rubin NVL72 platform combines 72 Rubin GPUs, 36 Vera CPUs, NVIDIA networking technologies, and data processing units to support large-scale AI computing.

CoreWeave provides cloud infrastructure focused on artificial intelligence workloads, including GPU computing, storage, and machine learning services for enterprise customers.

The company said its latest infrastructure developments are designed to keep data closer to AI computing resources and support larger AI models and agent-based applications.