Simulation-ready power and cooling infrastructure models
Vertiv™ is delivering converged physical infrastructure designs for the NVIDIA Vera Rubin DSX AI factory reference design and the NVIDIA Omniverse™ digital twin blueprint. Vertiv™ OneCore for NVIDIA Vera Rubin DSX enables AI factory builders with parameterized infrastructure models and deployment-ready building blocks that span power, cooling, controls, and lifecycle services featuring:
Vertiv™ OneCore for NVIDIA Vera Rubin DSX is designed to deliver:
faster with pre-engineered reference designs1
less time for infrastructure deployment2
more cumulative AI token output
in the first 12 months3
lower TCO4
1. Modeled. Time to an initial integrated design, configured and validated in the digital twin: 3 days vs. a 30-day conventional design process. Results vary by project
2. Modeled. On-site power and cooling deployment with Vertiv factory-built infrastructure: 45 days vs. a 90-day conventional on-site baseline. Results vary by project
3. Modeled. First 12 months measured from design kickoff: 317 vs. 245 productive days at equal installed capacity (3 + 45 vs. 30 + 90 days). Actual results will vary.
4. In aggregate, including the prefabricated shell, compared to traditional construction methods.
Orchestrating converged physical infrastructure for the AI era
Turn complex AI infrastructure from a collection of separate products into converged, simulation-ready physical system
Vertiv™ OneCore for NVIDIA Vera Rubin DSX
A groundbreaking, turnkey data center infrastructure platform designed to accelerate deployment and simplify delivery for hyperscale and large-scale environments.
- Standardized reference designs at 12.5, 50, 125, and 250 MW
- Scales seamlessly up to 1 GW
- Delivers clear, repeatable path from initial deployment to campus‑level expansion
