DigestAI news desk
Hardware & Compute updated 1 min read

CoreWeave shares rise after deploying first Nvidia Vera Rubin NVL72 cluster

CoreWeave Inc. shares increased by 4.5% following the announcement that the company has successfully deployed its first multi-rack Nvidia Vera Rubin NVL72 cluster. This infrastructure marks a significant milestone for the AI cloud provider, as it becomes the first to validate and bring online this specific configuration, which integrates hundreds of Nvidia Rubin GPUs designed specifically for…

2 sources

Key points

  • CoreWeave shares rose 4.5% after deploying the first validated Nvidia Vera Rubin NVL72 cluster for agentic AI.
  • Each rack in the new cluster contains 72 Rubin GPUs, 36 Vera CPUs, and Nvidia NVLink 6 connectivity.
  • New storage features include cross-region write acceleration and an Archive tier with 7 GB/s throughput per GPU.

In addition to the compute hardware, CoreWeave introduced enhanced storage capabilities to complement the new cluster. The company launched cross-region write acceleration, which allows data to be written at local latency while simultaneously migrating to a remote region in the background. Furthermore, a new Archive tier was introduced to provide lower-cost storage options without penalties for early deletion or reading. CoreWeave claims its AI Object Storage LOTA system achieves read speeds comparable to local NVMe drives, reducing latency by eightfold compared to traditional storage clusters and delivering up to 7 GB/s of throughput per GPU.

Full story from finance.yahoo.com · by Proactive · via Search: NVIDIA Open source ↗

CoreWeave shares climb as Nvidia Rubin cluster goes online

finance.yahoo.com · 16 September 2026

CoreWeave Inc (NASDAQ:CRWV) shares rose 4.5% Wednesday after the AI cloud provider deployed a multi-rack Nvidia Corp (NASDAQ:NVDA, XETRA:NVD) Vera Rubin NVL72 cluster and introduced new storage capabilities.

The deployment brings hundreds of Nvidia Rubin GPUs into a single cluster designed for agentic AI workloads, supporting training and inference jobs across multiple racks.

Each rack combines 72 Rubin GPUs and 36 Vera CPUs with Nvidia NVLink 6, Nvidia ConnectX-9 SuperNICs and Nvidia BlueField-4 data processing units.

CoreWeave said it was the first AI cloud provider to validate and bring up a Vera Rubin NVL72, using Nvidia Spectrum-X Ethernet networking to unify racks.

Alongside the deployment, the company introduced cross-region write acceleration, allowing data to be written at local latency while migrating to a second remote region in the background.

Its new Archive tier offers lower-cost storage without retrieval, early deletion or reading fees.

CoreWeave says its AI Object Storage LOTA delivers reads at local NVMe speeds and reduces latency eightfold compared with reading from a traditional storage cluster.

The company says its AI Object Storage system provides up to 7 GB/s of throughput per GPU.

This text was published by finance.yahoo.com and written by Proactive. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗

Coverage and discussion

2 sources
Topics · follow one to build your own front page
CoreWeaveNvidiaVera Rubin NVL72RubinVera

The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us.

Comments

via GitHub Discussions

More in Hardware & Compute

All →

Related stories