NVIDIA Unveils Omniverse DSX Blueprint for Gigawatt-Scale AI Factories

NVIDIA has announced the launch of its Omniverse DSX Blueprint, a groundbreaking open framework designed to facilitate the design and operation of gigawatt-scale AI factories. This initiative was unveiled by NVIDIA founder and CEO Jensen Huang during the GTC Washington, D.C. keynote, as reported by the NVIDIA Blog.

Innovative Approach to AI Factory Design

The Omniverse DSX Blueprint integrates design, simulation, and operations across AI factory facilities and leverages NVIDIA’s AI infrastructure stack. The framework unites ecosystem partners to set new standards for building and operating large-scale AI factories, validated at the AI Factory Research Center in Manassas, Virginia.

Digital Twin Technology and SimReady Assets

The blueprint employs digital twin technology, enabling partners like Jacobs to design and optimize AI factory layouts for maximum efficiency. Using SimReady OpenUSD assets from partners such as Siemens, Schneider Electric, and Vertiv, the process accelerates design decisions and reduces development risks.

Cadence’s Reality Digital Twin platform, built on the blueprint and powered by NVIDIA CUDA, allows engineers to simulate and optimize operations in a high-fidelity environment before construction, predicting potential failures and enhancing operational efficiency.

Optimizing Physical Build-Out and Operations

With partners like Bechtel and Vertiv, NVIDIA facilitates the delivery of prefabricated modules for AI factories, reducing build time and enabling modular scalability. Once operational, the digital twin serves as an operating system for monitoring and optimizing factory processes.

AI agents from partners like Phaidra and Emerald AI are integrated to optimize power consumption, enhancing grid flexibility and energy efficiency. This transforms the digital twins into self-learning systems that continuously improve factory operations.

Core Pillars of DSX

The DSX Blueprint is built on three pillars:

• DSX Flex – Dynamic Grid Collaboration: Enhances energy efficiency and grid cooperation using Omniverse telemetry and AI agents.
• DSX Boost – Performance-Per-Watt Optimization: Maximizes GPU throughput within existing power envelopes, improving energy efficiency by up to 30%.
• DSX Exchange – Unified IT/OT Integration: Provides a secure data and control fabric linking operational systems with NVIDIA’s software stack.

Significance and Future Impact

The DSX Blueprint promises scalability, efficiency, sustainability, and integration for AI factories. It supports a common architecture for facilities ranging from 100 megawatts to multi-gigawatts, enhancing token throughput and enabling renewable energy generation.

Together with partners like Eaton, GE Vernova, and Hitachi, NVIDIA combines expertise in power and cooling to deliver resilient operations optimized for NVIDIA platforms. The DSX ecosystem extends beyond hardware to include software and AI solutions, fostering dynamic simulation and autonomous control throughout the AI factory lifecycle.

As the demand for AI infrastructure grows, NVIDIA’s Omniverse DSX Blueprint is poised to accelerate the development and deployment of efficient, scalable AI factories across the globe. For more information, visit the NVIDIA Blog.

Image source: Shutterstock