Vertiv, a company focused on critical digital infrastructure, has announced an update to its production-grade digital twin capability, integrated into the NVIDIA Omniverse DSX Blueprint. The development is positioned as a step toward making AI factory infrastructure more configurable, simulation-ready, and adaptable to changing requirements.
As AI deployments continue to scale in size and capacity, there is increasing demand for data centres to convert successive generations of computing hardware into physical infrastructure more quickly and reliably. Traditional document-based workflows, along with separated processes across power, cooling, controls, and deployment teams, are described as less efficient in this context. Vertiv’s digital twin approach, referred to as SmartRun, uses a model-based method intended to allow infrastructure to be designed, simulated, and validated in an integrated way before physical construction.
By representing system configurations and dependencies in a virtual environment, the SmartRun digital twin is intended to help identify potential design changes earlier and reduce integration risks. It also enables simulation-based evaluation and supports coordination across different teams, with the aim of shortening the transition from planning to operational deployment.
The SmartRun digital twin initiative is described as the first phase of Vertiv’s broader roadmap for AI factory digital twins. These digital models are intended to connect compute development cycles with physical infrastructure readiness, covering stages from early configuration and simulation through deployment and lifecycle management.
In collaboration with NVIDIA, the initiative is intended to allow stakeholders to assess infrastructure options earlier and prepare for future generations of accelerated computing. Vertiv is expected to present SmartRun at Computex Taipei 2026, where it will be shown both as a physical infrastructure system and as a digital twin, demonstrating its use in design and simulation workflows.
The development also uses model-based systems engineering on the 3DEXPERIENCE platform, aiming to provide a unified environment for configuration, validation, and longer-term optimisation across the lifecycle of AI factory infrastructure.
