LIVINGSTON, N.J., July 3, 2025 /PRNewswire/ — , the AI Hyperscaler™, today announced a significant first in the AI cloud space: it is the inaugural AI cloud provider to deploy the latest NVIDIA GB300 NVL72 systems for its clients, with aggressive plans for worldwide scaling.
The NVIDIA GB300 NVL72 marks a revolutionary leap in performance, particularly for AI reasoning and agentic workloads. It promises an astounding 10x boost in user responsiveness, a 5x improvement in throughput per watt compared to the previous NVIDIA Hopper architecture, and a staggering 50x increase in output for reasoning model inference – a game-changer for complex AI tasks.
“At CoreWeave, we’re relentlessly pushing the envelope in AI development, providing the state-of-the-art cloud infrastructure essential for training the next generation of AI models,” said Peter Salanki, Co-Founder and Chief Technology Officer at CoreWeave. “Becoming the first to implement this groundbreaking platform is a testament to our commitment, and we are thrilled to empower innovators as they prepare for the next exciting era of AI.”
CoreWeave’s successful initial deployment of the NVIDIA GB300 NVL72 systems was a collaborative effort, with strategic partnerships with Dell, Switch, and Vertiv. This synergy has enabled unparalleled speed and efficiency in bringing these advanced NVIDIA GPUs onto CoreWeave’s robust AI cloud platform.
The integration of the GB300 NVL72 is seamless, deeply embedded within CoreWeave’s native cloud software ecosystem. This includes its CoreWeave Kubernetes Service (CKS) and Slurm on Kubernetes (SUNK), complemented by its advanced observability tools and a custom-designed Rack LifeCycle Controller (RLCC). Enhancing this ecosystem further, CoreWeave recently announced the direct integration of hardware-level data and cluster health events through Weights & Biases’ developer platform, a move that solidifies CoreWeave’s already strong position following its acquisition of the company earlier this year.
This latest achievement reinforces CoreWeave’s reputation for delivering prompt, market-leading access to the most advanced AI infrastructure, a critical requirement for leading AI research labs and enterprises globally. The introduction of the NVIDIA GB300s builds upon CoreWeave’s already extensive Blackwell fleet, which features the NVIDIA HGX B200 and the NVIDIA GB200 NVL72 system. This follows CoreWeave’s precedent-setting early access to NVIDIA H200 GPUs last year, and making NVIDIA GB200 NVL72 systems generally available as the first AI cloud provider.
Further underscoring its leadership, in June 2025, CoreWeave, in tandem with NVIDIA and IBM, submitted the most extensive MLPerf® Training v5.0 benchmark results ever recorded. Utilizing nearly 2,500 NVIDIA GB200 Grace Blackwell Superchips, the submission achieved a remarkable breakthrough, completing the complex Llama 3.1 405B model training in an astonishing 27.3 minutes. Notably, CoreWeave stands as the sole hyperscaler to attain the highest Platinum rating from SemiAnalysis’s GPU Cloud ClusterMAX™ Rating System, an independent benchmark for the AI cloud industry.
About CoreWeave
CoreWeave, the AI Hyperscaler™, operates a cutting-edge cloud platform engineered to power the next wave of artificial intelligence innovation. The company provides enterprises and premier AI laboratories with accelerated computing cloud solutions. Since 2017, CoreWeave has been steadily expanding its data center footprint across the United States and Europe. Recognized for its impact, CoreWeave was named one of TIME100’s most influential companies and secured a spot on Forbes’ Cloud 100 ranking in 2024. Discover more at www.coreweave.com.
Forward-Looking Statements
This press release contains forward-looking statements as defined by the Private Securities Litigation Reform Act of 1995. These statements, based on current expectations, forecasts, and assumptions, involve inherent risks and uncertainties that could cause actual results to materially differ from those discussed. Such factors, including general market, political, economic, and business conditions, are detailed in CoreWeave’s filings with the Securities and Exchange Commission, particularly within the sections “Special Note Regarding Forward-Looking Statements” and “Risk Factors” of CoreWeave’s Quarterly Report on Form 10-Q for the quarter ended March 31, 2025. All forward-looking statements herein reflect information available as of the date of this release, and CoreWeave assumes no obligation to update these statements due to new information or future events.
View original content to download multimedia:https://www.prnewswire.com/news-releases/coreweave-becomes-first-hyperscaler-to-deploy-nvidia-gb300-nvl72-platform-302497802.html
SOURCE CoreWeave
Original article, Author: Jam. If you wish to reprint this article, please indicate the source:https://aicnbc.com/3942.html
