Oct. 2, 2025 — AI and HPC infrastructure company VDURA, formerly called Panasas, today announced it has been selected by a Tier-1 U.S. federal system integrator to support a large defense GPU deployments.
While details on the installation were limited, VDURA said the program standardizes on the company’s V5000 Data Platform to accelerate AI and computational fluid dynamics (CFD) workloads, combining flash-class performance, unified tiering across NVMe and HDD, and built-in end-to-end encryptionto safeguard sensitive data, according to VDURA.
VDURA said the integrator chose VDURA’s V5000 Data Platform for its ability to deliver ultra-low latency, flash-class throughput, and seamless tiering under a single global namespace while meeting security, availability, and efficiency requirements.
Phase 1 of the deployment includes a 20 PB deployment of high-performance storage, achieving sustained transfer rates above 800 GB/s, with future scalability to 200 PB and 2.5 TB/s. The solution reduces cost-per-terabyte by over 60 percent compared to all-flash alternatives, improved energy efficiency by 44 percent, and lowered operational overhead to just half an FTE, VDURA said.
“This win validates our belief that true HPC+AI environments demand more than raw speed, they need architectural durability and scalability,” said Ken Claffey, CEO, VDURA. “By delivering flash-class responsiveness with object-storage economics, we’ve helped our customer launch a defense-grade AI platform that’s ready to scale and grow.”
Security & Compliance:
- End-to-end encryption built in to protect sensitive workloads throughout the pipeline.
- Unified global namespace with policy driven flash-to-disk tiering—no external data movers or manual migrations.
- Configurable erasure coding with up to 12-nines durability commitment to protect data integrity at scale.
