Dell Delivers World's First NVIDIA Rubin NVL72 AI Supercomputer to CoreWeave

In a landmark development for the artificial intelligence industry, Dell has successfully delivered the world's first operational NVIDIA Rubin NVL72 AI supercomputer system to CoreWeave. According to social media updates from both companies, the groundbreaking system has successfully passed all its validation tests, marking its readiness for deployment.

At the heart of this delivery is Dell's custom-built PowerEdge XE9812 liquid-cooled server, specifically engineered around the NVIDIA Rubin NVL72 architecture. This powerful server integrates an impressive 72 next-generation Rubin GPUs and 36 Vera CPUs, creating a tightly coupled system designed for maximum performance and efficiency in demanding AI workloads.

The capabilities of this new supercomputer are tailored for the next wave of AI development. It is optimized for training ultra-large-scale AI models with trillion-level parameters, as well as complex Mixture of Experts (MoE) models. Furthermore, the system is engineered to deliver a lower cost-per-token during large-scale AI inference, making advanced AI more economically viable.

Looking ahead, this delivery is a crucial step in CoreWeave's strategy. As a leading "Neocloud" enterprise, CoreWeave plans to integrate these NVIDIA Rubin-based systems into its AI cloud platform starting in the second half of 2026. This milestone delivery solidifies the partnership between Dell, NVIDIA, and CoreWeave, paving the way for unprecedented computational power in the cloud.