CoreWeave, Inc. (CRWV) is enhancing its capabilities for AI workloads by deploying multi-rack NVIDIA Vera Rubin NVL72 clusters on CoreWeave Cloud. This expansion aims to support compute-intensive AI workloads while ensuring data accessibility. The NVL72 configuration involves 72 Rubin GPUs and 36 Vera CPUs per rack, facilitating hundreds of GPUs to function as a single scale-out cluster.
The new AI Object Storage capabilities include cross-region write acceleration, which allows local data writes with background replication, and a lower-cost Archive tier for infrequently accessed datasets. CoreWeave’s data-transport accelerator can minimize latency by up to 8x, achieving throughput of 7 GB/s per GPU.
CoreWeave is competing in a crowded AI infrastructure market, facing rivals like Microsoft and Nebius Group N.V. (NBIS). Notably, NBIS plans to invest £1.7 billion in UK AI infrastructure, aiming to deploy over 5 GW of AI computing capacity by 2030, while Microsoft is expanding its Azure infrastructure to meet growing AI demands.
5 Stocks Our Experts Predict Could Double In the Next Year
By submitting your email, you'll also get a free pivot & flow membership. A free daily market overview. You can unsubscribe at any time.
