hitachi-digital-twin-03-campus
Image credit: Hitachi Energy

Hitachi Energy has developed a power system simulation capability integrating NVIDIA Omniverse libraries to help data centre engineers evaluate designs for AI infrastructure before construction.

The initial application focuses on emerging 800 V DC architectures intended to support higher power densities in next-generation AI data centres. The simulation combines AI load scenarios with Hitachi Energy’s power system models, allowing teams to compare design alternatives and assess system resilience.

Using OpenUSD-based workflows, engineers can evaluate operating conditions, load profiles and resilience scenarios across the power supply chain from grid to rack. According to Hitachi Energy, the capability is intended to help validate design assumptions and identify potential performance issues earlier in planning.

The simulation interface also incorporates an AI assistant that allows users to explore the digital twin through natural-language queries, retrieve results and compare scenarios.

Marco Berardi, Head of Grid & Power Quality Solutions and Service at Hitachi Energy, said: “As AI factories scale, access to electricity, power system design and the ability to deploy infrastructure at speed are becoming increasingly important.”

The development builds on Hitachi Energy’s collaboration with NVIDIA on 800 V DC infrastructure and forms part of the NVIDIA DSX Sim partner ecosystem.

It also follows the qualification of Hitachi Energy’s battery energy storage system as NVIDIA DSX Ready BESS. According to the company, that qualification identifies equipment meeting applicable NVIDIA DSX reference design requirements for AI power infrastructure.

Source: Hitachi Energy

🌍 Explore related coverage