NVIDIA GTC 2025: NVIDIA-Certified Program Expands to Enterprise Storage for Faster AI Factory Deployment

By Matthew Hausmann, NVIDIA Corp.

AI deployments thrive on speed, data and scale.

(Image source: NVIDIA Corp.)

That’s why NVIDIA Corp. is expanding NVIDIA-Certified Systems to include enterprise storage certification – for streamlined AI factory deployments in the enterprise with accelerated computing, networking, software and storage.

As enterprises build AI factories, access to high-quality data is imperative to ensure optimal performance and reliability for AI models. The NVIDIA-Certified Storage program announced today at the NVIDIA GTC 2025 AI conference validates that enterprise storage systems meet stringent performance and scalability data requirements for AI and HPC workloads.

Leading enterprise data platform and storage providers are already onboard, ensuring businesses have trusted options from day one. These include DDN, Dell Technologies, Hewlett Packard Enterprise, Hitachi Vantara, IBM, NetApp, Nutanix, Pure Storage, VAST Data and WEKA.

Building Blocks for New Class of Enterprise Infrastructure
At GTC, NVIDIA also announced the NVIDIA AI Data Platform, a customizable reference design to build a new class of enterprise infrastructure for demanding agentic AI workloads.

The NVIDIA-Certified Storage designation is a prerequisite for partners developing agentic AI infrastructure solutions built on the NVIDIA AI Data Platform. Each of these NVIDIA-Certified Storage partners will deliver customized AI data platforms, in collaboration with NVIDIA, that can harness enterprise data to reason and respond to complex queries.

NVIDIA-Certified was created more than 4 years ago as the industry’s 1st certification program dedicated to tuning and optimizing AI systems to ensure optimal performance, manageability and scalability. Each NVIDIA-Certified system is rigorously tested and validated to deliver enterprise-grade AI performance.

There are now 50+ partners providing 500+ NVIDIA-Certified systems, helping enterprises reduce time, cost and complexity by giving them a wide selection of performance-optimized systems to power their accelerated computing workloads.

NVIDIA Enterprise Reference Architectures (RAs) were introduced last fall to provide partners with AI infrastructure best practices and configuration guidance for deploying NVIDIA-Certified servers, NVIDIA Spectrum-X networking and NVIDIA AI Enterprise software.

Solutions based on NVIDIA Enterprise RAs are available from the world’s leading systems providers to reduce the time, cost and complexity of enterprise AI deployments. Enterprise RAs are now available for a wide range of NVIDIA Hopper and NVIDIA Blackwell platforms, including NVIDIA HGX B200 systems and the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU.

These firm’s technologies and partner solutions are the building blocks for enterprise AI factories, representing a new class of enterprise infrastructure for high-performance AI deployments at scale.

Enterprise AI Needs Scalable Storage
As the pace of AI innovation and adoption accelerates, secure and reliable access to high-quality enterprise data is becoming more important than ever. Data is the fuel for the AI factory. With enterprise data creation projected to reach 317ZB annually by 2028 ^(*), AI workloads require storage architectures built to handle massive, unstructured and multimodal datasets.

NVIDIA’s expanded storage certification program is designed to meet this need and help enterprises build AI factories with a foundation of high-performance, reliable data storage solutions. The program includes performance testing as well as  validation that partner storage systems adhere to design best practices, optimizing performance and scalability for enterprise AI workloads.

NVIDIA-Certified Storage will be incorporated into the company’s Enterprise RAs, providing enterprise-grade data storage for AI factory deployments with full-stack solutions from global systems partners.

Certified Storage for Every Deployment
This certification builds on existing NVIDIA DGX systems and NVIDIA Cloud Partner (NCP) storage programs, expanding the data ecosystem for AI infrastructure.

Storage certification programs are aligned with their deployment models and architectures:

• NVIDIA DGX BasePOD and DGX SuperPOD Storage Certification – designed for enterprise AI factory deployments with NVIDIA DGX systems.

• NCP Storage Certification – designed for large-scale NCP Reference Architecture AI factory deployments with cloud providers.

• NVIDIA-Certified Storage Certification – designed for enterprise AI factory deployments with NVIDIA-Certified servers available from global partners, based on NVIDIA Enterprise RA guidelines.

With this framework, organizations of all sizes – from cloud hyperscalers to enterprises – can build AI factories that process massive amounts of data, train models faster and drive more accurate, reliable AI outcomes.

^(*) Source: IDC, Worldwide IDC Global DataSphere Forecast, 2024–2028: AI Everywhere, But Upsurge in Data Will Take Time, doc #US52076424, May 2024

Resources:
More about how NVIDIA-Certified Systems deliver high-speed performance and attend these related sessions at GTC:
Unlock the Power of NVIDIA Certified Systems: Live Q&A With Experts    
Everything You Want to Ask About NVIDIA Enterprise Reference Architectures