Naval Postgraduate School Deploys NVIDIA AI via Vertiv Infrastructure

The project centers on the Vertiv SmartIT rack-level architecture, a modular system designed to manage the extreme thermal and electrical demands of modern accelerated computing. By integrating power distribution, liquid cooling, and fluid management into a single pre-engineered unit, the setup allows the school to host heavy AI workloads while maintaining operational stability. Vertiv CEO Gio Albertazzi noted that the deployment serves as a blueprint for organizations looking to scale AI capacity within current physical constraints.

The installation, which involved a three-rack configuration, highlights the shift toward treating data center infrastructure as a unified, factory-integrated product rather than a collection of disparate parts. Charlie Boyle, vice president of DGX systems at NVIDIA, emphasized that the collaboration provides the necessary foundation for faculty and researchers to pursue complex AI research. This approach reduces the integration risks typically associated with high-density computing upgrades, establishing a scalable model for future institutional growth.