NVIDIA, Emerald Team on Flexible AI Factories
Six Large Utilities Support Plan for AI Software to Manage Power During Grid Peaks
The power-flexible AI factory is coming soon.
Nvidia and energy management software provider Emerald AI announced a sweeping consortium with six large U.S. power providers to deploy data centers that act as active grid assets rather than passive consumers.
By combining onsite power generation with compute-aware orchestration software, the alliance aims to drastically shorten the time it takes to bring new AI capacity online while providing much-needed relief to an increasingly strained U.S. electrical grid.
To make this strategy work at scale, the initiative brings together a heavyweight roster of energy providers committed to the new hybrid framework:
AES
Constellation
Invenergy
NextEra Energy
Nscale Energy & Power
Vistra
“AI factories are the engines of the intelligence era, and like any great engine, every system must be designed together - energy, compute, networking and cooling as one architecture,” said Jensen Huang, founder and CEO of Nvidia. “Nvidia and Emerald AI are working together to enable a future for AI where performance, efficiency and grid responsiveness can be tapped into immediately.”
The Mechanics of Flexibility
How does a massive data center transform into a grid-responsive asset? The answer lies in the tight orchestration of both silicon and software.
The next-generation AI factories will be built on Nvidia’s Vera Rubin DSX reference design, which incorporates a new software library known as DSX Flex, which enables the physical infrastructure to communicate with power grid services.
The key component is Emerald AI’s Conductor platform, which choreographs the IT load with energy resources, such as onsite natural gas generation or battery storage. During moments of peak grid stress, Conductor can throttle non-essential compute or island the facility using onsite generation, effectively freeing up capacity for the local utility.
“AI factories are too valuable to be treated as either passive loads or permanent islands,” said Varun Sivaram, founder and CEO of Emerald AI. “Emerald Conductor orchestrates compute flexibility alongside onsite energy resources to support the grid, so projects can connect sooner, preserve quality of service for AI tenants and ultimately strengthen the power system around them.”
Over the last year, Emerald AI and NVIDIA have run AI power flexibility demonstrations at five commercial data centers around the world.
DSX Flex is expected to be deployed at commercial scale later this year at the NVIDIA AI Factory Research Center in Virginia, planned as one of the world’s first power-flexible AI factories with NVIDIA Vera Rubin infrastructure.
