NVIDIA's DGX GB300 supercomputer powers naval AI research
1,500 students and 600 faculty get on-premises access to massive AI compute.
NVIDIA CEO Jensen Huang officially commissioned a DGX GB300 supercomputer at the Naval Postgraduate School (NPS) in Monterey, California, on July 23, 2026. This powerful AI platform is designed to provide on-premises, large-scale computing for model training and inference. It will serve over 1,500 students and 600 faculty members, enabling them to work on computationally intensive projects without relying on cloud resources. The DGX GB300 system is tailored to support applications ranging from weather prediction and cybersecurity to disaster response, addressing both academic research and real-world defense needs.
The commissioning underscores NVIDIA's ongoing collaboration with the U.S. military to accelerate AI capabilities. By hosting the supercomputer on campus, NPS gains low-latency access to cutting-edge hardware for hands-on education and classified research. This initiative is expected to train a new generation of naval officers and scientists in AI, directly impacting operational readiness and national security. The DGX GB300's architecture likely leverages NVIDIA's latest GPU and networking technologies, though specific performance metrics were not disclosed.
- DGX GB300 commissioned at Naval Postgraduate School by Jensen Huang on July 23, 2026.
- Serves 1,500+ students and 600 faculty with on-premises AI compute for training and inference.
- Target applications: weather prediction, cybersecurity, and disaster response.
Why It Matters
Militarizes AI education with dedicated on-premises supercomputing, enabling classified research and operational readiness.