DoD tests AI models that make it easy to switch from vendors like Palantir
Investing.com -- CoreWeave Inc (NASDAQ:CORE) stock climbed 3.2% Thursday after the AI cloud provider announced it has become the first to deploy NVIDIA’s (NASDAQ:NVDA) latest GB300 NVL72 platform, setting a new benchmark for AI reasoning and performance.
The GB300 NVL72 represents a significant advancement in AI infrastructure, delivering up to 10x boost in user responsiveness and 5x improvement in throughput per watt compared to the previous NVIDIA Hopper architecture. This translates to a 50x increase in output for reasoning model inference, according to CoreWeave.
The new platform features NVIDIA Blackwell Ultra GPUs with 1.5x more AI compute FLOPS than standard Blackwell GPUs, expanded HBM3e memory with up to 21TB of high-bandwidth GPU memory per rack, and fifth-generation NVIDIA NVLink providing 130TB/s of aggregate bandwidth.
CoreWeave has optimized its cloud platform specifically for the GB300 NVL72, enhancing its Kubernetes service, Slurm on Kubernetes offering, and introducing new monitoring capabilities including Cabinet Wrangler and Cabinet Details dashboard for rack-scale insights.
The deployment was completed in collaboration with Dell Technologies (NYSE:DELL), Switch, and Vertiv, demonstrating CoreWeave’s ability to rapidly bring cutting-edge AI infrastructure to market.
"The bringup of the GB300 on CoreWeave is incredibly exciting. This platform and CoreWeave’s ability to quickly bring it to market via its AI cloud platform, are a major enabler for AI-driven filmmaking to become a reality for our industry," said Naeem Talukdar, Co-Founder and CEO of Moonvalley.
Earlier this year, CoreWeave published benchmarks showing its GB200-based instances delivered 2.86x better per-GPU inference performance compared to H100-based instances, highlighting the company’s focus on optimizing AI workload performance.
This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.