NVIDIA commissions DGX GB300 supercomputer for Naval Postgraduate School in Monterey

NVIDIA brought a new AI supercomputer online for the U.S. Naval Postgraduate School (NPS) in Monterey, California, commissioning an NVIDIA DGX GB300 system that will support graduate instruction and research workloads on campus. NVIDIA founder and CEO Jensen Huang visited the school during NPS’s three-day Converge @ NPS event, according to an announcement from NVIDIA.

The DGX GB300 system is intended to give NPS’s on-campus community large-scale AI compute capacity, including capabilities for model training and inference. NVIDIA said the platform is designed to serve NPS’s more than 1,500 in-resident students and roughly 600 faculty, who use the school for applied research tied to real-world defense and operational challenges.

NVIDIA also linked the new commissioning to an AI Technology Center on the Monterey campus. The company said the DGX GB300 system is anchored inside that center, which is described as a dedicated hub for AI research and graduate instruction, with the new hardware functioning as a core component of the center’s computational resources.

Huang framed the effort around the importance of timely information in high-stakes environments, telling attendees that “nothing is more valuable to you than information and insight,” and adding that he could not imagine anything more important. NPS leaders also emphasized that advanced computing access is part of how military education can keep pace with evolving technology, including the responsibilities that come with using it.

NVIDIA said the DGX GB300 is paired with NVIDIA Mission Control software, and positioned the system as enabling work across foundation model development and simulation. The announcement specifically cited in-house foundation model training, large-scale high-fidelity simulations, and the development of AI tools for applications spanning weather prediction, cybersecurity, and disaster resilience and response planning.

Beyond the hardware, NVIDIA said the collaboration includes curriculum support through its Deep Learning Institute, which provides instructor toolkits meant to help faculty weave AI into graduate programs across departments. NPS’s graduate model, as described by NVIDIA, centers applied research with direct relevance to real-world problems, and the company said NPS’s “regular cadence” of hackathons has already produced advances in areas such as autonomy, ocean research, and operations planning that require substantial computing capacity.

The announcement highlighted example research directions that the DGX GB300 is expected to power. NVIDIA said researchers are using AI to model sea conditions, predict atmospheric changes, and build digital twins of complex environments. In that context, a digital twin is a computational representation of a real-world system that can be used for simulation, scenario testing, and decision support under changing or uncertain conditions.

NVIDIA further described a partnership with MITRE, a nonprofit organization, connected to digital twin frameworks using NVIDIA Omniverse libraries. NVIDIA said NPS developed a framework that supports high-fidelity digital twin environments for simulating real-world navigation and decision-making under uncertain conditions, and that NPS and NVIDIA researchers have presented related AI modeling and simulation work at NVIDIA GTC Washington, D.C.

Hardware integration for the project involved additional partners named by NVIDIA. The company said DDN contributed high-performance data infrastructure, VAST Data provided a unified data platform for secure data access and management across edge, core, and cloud environments, and Vertiv supplied the racks, cooling, and power infrastructure, including installation sequencing, testing and commissioning support, and fluid management and monitoring. NVIDIA did not provide the system’s exact specifications or throughput figures in the announcement, so it remains unclear how the NPS deployment’s performance compares with other DGX GB300 installations.