Vertiv has announced new converged physical infrastructure innovations designed to support the next generation of AI factories built on NVIDIA’s Vera Rubin DSX architecture, addressing the growing complexity and scale of AI deployments.
As AI factories expand in density and power requirements, organisations are facing increasing pressure to reduce deployment timelines, optimise infrastructure utilisation, and minimise integration risks. Vertiv’s approach introduces simulation-ready infrastructure models, enabling operators to validate and optimise power and cooling systems before physical deployment begins.
At the core of this initiative is Vertiv’s converged physical infrastructure model, which integrates power, cooling, controls, and lifecycle services into a unified, system-level design. This model is built on repeatable infrastructure blocks, validated interfaces, and digital continuity, allowing for more scalable and predictable AI factory deployments.
A key component is the Vertiv™ OneCore modular architecture, based on standardised 12.5MW infrastructure blocks that can be scaled from smaller AI clusters to gigawatt-scale AI factories. This approach simplifies expansion while improving consistency, coordination, and operational performance across deployments.
According to Scott Armul, the shift towards AI factories requires a fundamental rethink of how digital infrastructure is designed and deployed, moving from isolated systems to integrated, simulation-driven environments that accelerate time to deployment and improve efficiency.
The collaboration also leverages digital twin simulation and real-time system modelling, enabling enterprises to reduce field integration risks, accelerate operational readiness, and optimise performance across the entire infrastructure stack, from grid connectivity to chip-level thermal management.
Vladimir Troy highlighted that integrating simulation-ready infrastructure into NVIDIA’s AI factory designs provides the validated building blocks needed to scale deployments with greater confidence and speed.
With this development, Vertiv is positioning its converged infrastructure approach as a foundational layer for future AI factory deployments across hyperscale, colocation, enterprise, and emerging AI environments, enabling organisations to build and scale high-performance AI infrastructure with improved efficiency, reliability, and operational control.