Vertiv Accelerates Deployment of Converged Physical Infrastructure for NVIDIA Vera Rubin DSX AI Factories

Dubai, UAE: Vertiv (NYSE: VRT), a prominent player in critical digital infrastructure, has announced its involvement in enhancing converged physical infrastructure designs specifically for the NVIDIA Vera Rubin DSX AI factory reference design and the NVIDIA Omniverse DSX Blueprint. This collaboration marks a significant step in addressing the growing complexities and demands of AI factories.

The Need for Advanced Infrastructure Solutions

As AI factories evolve, they are experiencing increased density, complexity, and power requirements. Operators face mounting pressure to expedite deployment timelines, optimize infrastructure utilization, and mitigate integration risks. In response, Vertiv has introduced a new infrastructure design methodology aimed at simplifying these challenges. This approach not only enhances confidence prior to construction but also accelerates the time to operational capacity.

Through its partnership with NVIDIA, Vertiv is delivering simulation-ready digital power and cooling assets, validated interfaces, and repeatable infrastructure building blocks. These innovations are designed to facilitate quicker and more assured deployments of AI factories.

A System-Level Model for Integration

Vertiv’s approach to converged physical infrastructure represents an expansion of its existing methodologies. This system-level model integrates power, cooling, controls, and services into cohesive designs that are optimized across the entire power and thermal chain. The framework is built upon five foundational elements: repeatable building blocks, defined interfaces, system orchestration, digital continuity, and lifecycle support. Collectively, these components aim to enhance the scalability of AI factory operations by minimizing design complexity and improving coordination across various infrastructure domains.

At the heart of this initiative is a scalable building block architecture centered around standardized 12.5MW infrastructure blocks from Vertiv OneCore integrated modular solutions. These blocks can be combined and configured to accommodate a range of deployments, from smaller AI clusters to large-scale gigawatt AI factories. By establishing validated interfaces and repeatable designs, Vertiv seeks to streamline scaling while enhancing deployment consistency and operational performance.

Transforming AI Infrastructure Design

Scott Armul, Chief Product and Technology Officer at Vertiv, emphasized the transformative nature of AI factories on digital infrastructure design. He noted that the goal is to transition complex AI infrastructure from disparate products into unified, simulation-ready physical systems. This collaboration with NVIDIA enables customers to accelerate their journey from design to deployment. By integrating Vertiv’s power and cooling solutions with validated interfaces and digital models, customers can enhance development speed, operational confidence, and overall efficiency.

Vertiv’s collaboration is pivotal in developing digitally validated AI factory infrastructure. This is achieved through real-time simulation and system-level modeling prior to physical deployment. The benefits of this approach include:

• Reduced deployment complexity and field integration risks.
• Accelerated time to operational readiness.
• Improved coordination across power, cooling, and controls.
• Optimized performance from grid connection through chip-level thermal management and heat-reuse pathways.

Comprehensive Infrastructure Solutions

Vertiv’s contribution is anchored in its extensive portfolio of critical power, thermal management, integrated controls, and lifecycle services. This cohesive converged physical infrastructure aims to deliver both rapid deployment and cumulative system-level advantages. Unlike traditional modular or prefabricated solutions that primarily focus on scheduling, Vertiv’s approach seeks to enhance performance, efficiency, and reliability through standardized interfaces and repeatable building blocks.

Vladimir Troy, Vice President of AI Infrastructure at NVIDIA, highlighted the necessity for a converged approach to physical infrastructure as AI factories scale to unprecedented levels of power and density. He stated that by incorporating simulation-ready infrastructure models into the NVIDIA Vera Rubin DSX design, Vertiv is providing essential building blocks and validated interfaces that expedite the transition from design to operational readiness.

Future Directions in AI Infrastructure

The collaborative output, Vertiv OneCore Rubin DSX, is a design outcome rooted in converged physical infrastructure. Vertiv plans to continue refining this model for future compute generations, providing AI factory builders with parameterized infrastructure models and deployment-ready building blocks that encompass power, cooling, controls, and lifecycle services.

This initiative is expected to inform future converged infrastructure offerings across various sectors, including hyperscale, colocation, enterprise, and emerging AI deployment environments.

For more information on Vertiv solutions for AI infrastructure, visit Vertiv.com.

About Vertiv: Vertiv (NYSE: VRT) integrates hardware, software, analytics, and ongoing services to ensure that critical applications operate continuously and efficiently. The company addresses the pressing challenges faced by data centers, communication networks, and commercial facilities with a comprehensive portfolio of power, cooling, and IT infrastructure solutions. Headquartered in Westerville, Ohio, Vertiv operates in over 130 countries.

Follow the latest cybersecurity developments, threat intelligence and breaking updates from across the Middle East.