Hitachi Unveils 800 VDC Power Supply Simulation, Enabled by the NVIDIA Omniverse DSX Blueprint, to Accelerate Gigawatt-Scale AI Factories
-
Hitachi ’s power solution for AI infrastructure integrates with NVIDIA Omniverse DSX Blueprint as SimReady assets to accelerate development ofAI Factory digital twins. -
When used in digital twin design phases, the
Hitachi SimReady assets enable simulation to accelerate delivery of quality, modular power needed for gigawatt-scale AI factories. -
Innovation combines
Hitachi 's industry-leading expertise in physical AI, power, automation, and control architecture.
The three-dimensional (3D) simulation model of Hitachi’s solution, enabled by OpenUSD, replicates power behavior across the full electrical chain—from the utility grid to the data center rack—helping data center developers and utilities design, simulate, and deploy AI factories faster and with confidence. GTC 2026 attendees can experience the new grid-to-rack architecture demo in Hitachi’s booth #344 March 16 to 19.
Bridging the Physical and Digital Worlds
The simulation model, aggregated using OpenUSD and powered by industrial software providers and Omniverse libraries, equips data center developers and utilities with the following key features and benefits:
- SimReady Asset Integration: The power and control architecture integrates with the NVIDIA Omniverse DSX Blueprint as SimReady assets, bridging the gap between design and operational workflows.
- Grid Stability: The power solution design demonstrates stable, consistent power delivery to 800 VDC racks while maintaining grid compliance across a wide range of operating use cases. In the 3D simulation, Hitachi’s advanced power electronics and digital control algorithms analyze power quality disturbances caused by spiky AI workloads and smooths the power, leveraging battery energy storage systems as needed.
- Predictive Management: The design can integrate existing thermal, condition and health models of assets which enables predictive and prescriptive operations and maintenance from the data center to the substation and grid.
Hitachi’s latest development is the next step in its collaboration with NVIDIA to bring to market efficient, reliable DC power delivery. Rapidly growing AI workloads are forecasted to increase electricity demand up to 125 gigawatts of capacity by 2030, driving major power system expansion and modernization. An 800 VDC architecture is instrumental in feeding the increased power to data centers as it can manage 15x more power than legacy systems—enabling greater compute density with lower electricity use in a smaller footprint.
“No one has yet experienced the AI workloads and power demands of tomorrow’s gigawatt-scale AI factories. The aim is to scale fast without compromising quality and to safeguard the grid, chips, and entire electrical chain,” said
“As energy becomes the primary bottleneck for AI infrastructure, the industry requires sophisticated simulation tools to manage the unprecedented power densities and interconnection backlogs of gigawatt-scale AI factories,” said
To learn more about this and other
Trademark Notice: All trademarks and product names are the property of their respective owners.
About
Through its Social Innovation Business (SIB) that brings together IT, OT (Operational Technology) and products,
View source version on businesswire.com: https://www.businesswire.com/news/home/20260317248989/en/
Media Contact:
heather@211comms.com
Source: