NVIDIA Deploys DGX GB300 AI Supercomputer at Naval Postgraduate School
NVIDIA AI Supercomputer Comes Online at Naval Postgraduate School
On, NVIDIA founder and CEO Jensen Huang visited the Naval Postgraduate School in Monterey, California, to commission an NVIDIA DGX GB300 system.
The Tech TL;DR:
- Hardware Deployed: An NVIDIA DGX GB300 supercomputer running NVIDIA Mission Control software is now live on the Monterey campus.
- Operational Scope: Provides model training and inference for over 1,500 students and 600 faculty handling weather prediction, cybersecurity, and disaster response.
- Ecosystem Integration: Built with infrastructure support from DDN, VAST Data, Vertiv, and frameworks built on NVIDIA Omniverse libraries in partnership with MITRE.
Architectural Specifications and On-Premises Capabilities
The newly commissioned DGX GB300 system integrates tightly with existing defense research pipelines, moving computing overhead away from cloud dependencies and into a secured on-premises environment. Per details shared during the Converge @ NPS event, the supercomputer supports intensive model training and low-latency inference workloads required for complex operational environments. Naval Postgraduate School researchers are utilizing these hardware capabilities to model sea conditions, forecast atmospheric changes, and build high-fidelity digital twins of complex operational environments.
To construct these digital twin environments for navigation and decision-making under uncertainty, NPS leverages frameworks built on NVIDIA Omniverse libraries in partnership with nonprofit organization MITRE. Managing the massive input/output (I/O) demands of these high-performance computing (HPC) workflows required a multi-vendor hardware ecosystem. DDN contributed high-performance data infrastructure for scaling AI datasets, while VAST Data supplied a unified data platform to streamline secure access across edge, core, and cloud vectors. Physical deployment, power distribution, fluid management, and thermal monitoring were engineered and installed by Vertiv.
Curriculum Modernization and Developer Enablement
Beyond hardware installations, the collaboration embeds artificial intelligence directly into graduate coursework. Through an expansion with the NVIDIA Deep Learning Institute, NPS faculty are integrating instructor toolkits across space operations, ocean science, and cyber departments.
The pedagogical approach emphasizes first-principles software engagement over abstract theory.
# Example deployment verification for local model training nodes
curl -s --unix-socket /var/run/docker.sock http://localhost/v1.41/containers/json | jq '.[].Names'
nvidia-smi --query-gpu=index,name,utilization.gpu,memory.used --format=csv
Strategic Integration and Future Horizons
The DGX GB300 deployment anchors the existing NVIDIA AI Technology Center on the Monterey campus, a dedicated hub for graduate research and instruction. Researchers and engineers seeking deeper technical documentation on these implementations can reference whitepapers available via GitHub developer repositories or explore hardware documentation on NVIDIA Developer portals.