Article

Global AI has completed the deployment of the largest NVIDIA GB300 NVL72 cluster in New York's Endicott facility, marking a significant expansion of its sovereign AI infrastructure. The company plans to install NVIDIA Vera Rubin NVL72 systems across its U.S. data centers to enhance its model training and inference capabilities.

This expansion will support a critical IT load of 16 MW, with plans to ramp up to 100 MW throughout 2026. The Vera Rubin platform aims to deliver higher performance and lower costs for AI workloads, enhancing Global AI's compute capacity growth roadmap.