White Paper

NCP High-Performance Storage Reference Design for NVIDIA GB200 and GB300 NVL72 Systems

When you build and operate a large-scale GPU cloud, you need storage that can keep up and never slow you down.

This reference architecture details the validated VAST Data high-performance storage (HPS) reference design for NVIDIA GB200 and GB300 NVL72 systems, built specifically for NVIDIA Cloud Partners (NCP). Inside, you'll see exactly how VAST's AI Operating System, and the industry's first Disaggregated, Shared-Everything (DASE) architecture, with flexible CNodes and DBoxes, delivers the throughput, scale, reliability, and multi-tenant security that hyperscale AI infrastructure demands.

What you'll learn:

Validated configurations for deployments from 1,152 to 41,472 GPUs
How VAST exceeds NVIDIA's HPS performance requirements
How to consolidate storage silos and reduce TCO without sacrificing performance
Zero Trust multi-tenancy built for cloud service providers

Download the reference design and see how the world's largest GPU cloud deployments are built on VAST Data.

Want to receive info about products and services from VAST? You may unsubscribe from these communications at anytime.

By submitting this form, you consent that the information you provide will be collected and processed in accordance with our Privacy Policy, and may be shared with our partners to help fulfill and follow up on this request.

* Required field.