NVIDIA and Emerald AI are working with AES, Constellation, Invenergy, NextEra Energy, Nscale Energy & Power, and Vistra to develop AI factories that connect to the power grid faster and operate as flexible energy assets. These systems support AI workloads and generate AI tokens and intelligence. The collaboration brings together technology, energy, and infrastructure companies to support AI development and grid reliability in the United States.
NVIDIA’s Vera Rubin DSX AI Factory design and DSX Flex software connect AI factories to power grid services and support multiple deployment configurations. Factories can use co-located generation and storage as bridge power and later supply energy to the grid during operation. This approach supports faster deployment of AI capacity and integration with power systems.
Emerald AI’s Conductor platform coordinates compute demand with onsite generation, batteries, and other behind-the-meter resources to manage power use and maintain service levels for AI workloads. The DSX architecture also supports deployments without co-located energy resources, enabling larger and faster grid connections while reducing infrastructure requirements.
“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.”
“AI factories are too valuable to be treated as either passive loads or permanent islands,” said Varun Sivaram, founder and CEO of Emerald AI. “They produce tremendously valuable AI tokens and knowledge, and with DSX Flex, they can also provide measurable relief back to the grid. 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.”
Building AI Factories That Strengthen the Grid
Power-flexible AI factories can unlock up to 100 gigawatts of capacity across the U.S. power grid by combining infrastructure design, use of existing assets, and new generation where required. These systems can adjust power use during periods of grid stress, reducing the need for additional grid expansion.
AI factories convert electricity into AI tokens, models, and intelligence. This requires changes in computing systems and in the planning, construction, and operation of energy infrastructure.
Gigawatt-scale AI projects are using co-located generation and storage to address delays in grid interconnection. Isolating these resources can reduce asset utilization, increase cost per AI token, and limit their role in grid operations.
AES, Constellation, Invenergy, NextEra Energy, Nscale Energy & Power, and Vistra are building generation capacity to meet AI-driven demand. The companies will evaluate generation models for AI factories based on the NVIDIA and Emerald AI architecture, including hybrid systems that combine grid and co-located power. This approach supports faster power availability and enables grid-connected AI factories to use co-located resources where available.
“Grid flexibility will be key to addressing AI’s unprecedented demand while supporting system reliability,” said Andrés Gluski, CEO of AES. “At AES, we are enabling next-generation AI infrastructure to accelerate our clients’ time to power. DSX Flex embeds flexibility from the outset, allowing AI infrastructure to operate as a grid asset that supports faster, more efficient growth.”
“As the largest producer of clean energy in the U.S., we know data centers have enormous potential to unlock energy infrastructure investment, job creation and benefits for our communities,” said Joe Dominguez, president and CEO of Constellation. “They can also address the need for additional capacity through demand response. We don’t have a supply problem – we have a peak problem. By effectively using what we already have, including power-flexible AI factories that also enable AI-powered demand response, we can accommodate new load growth more efficiently.”
“AI is changing how we’re thinking about energy, and our customers need power fast, with the ability to scale over time,” said Michael Polsky, founder and CEO of Invenergy. “Combining near-term generation solutions with a path to full grid connection and flexible operations is an innovative and efficient way to help our customers meet their energy needs faster while keeping the system reliable.”
“To meet unprecedented new electricity demand while maintaining a reliable and resilient grid, now more than ever, we need to add generation resources,” said John Ketchum, chairman, president and CEO of NextEra Energy. “We also need technologies that allow new demand and related generation to integrate into the grid quickly and at the lowest possible cost. NextEra looks forward to working with NVIDIA and Emerald AI to help design efficient energy campuses and flexible AI factories that economically support rising demand while further strengthening America’s energy infrastructure.”
“We are committed to stabilizing the grid and helping West Virginia families and businesses have ready access to the power they need,” said Daniel Shapiro, chief power and energy officer of Nscale Energy & Power. “When we’re interconnected, we’ll be there on the grid’s highest-demand days to supply electricity back – that’s what 2 gigawatts scaling to 8 gigawatts of onsite generation means. Nscale’s Monarch campus is a power asset for West Virginia, not a load on it.”
“U.S. grids are designed to handle the highest-peak demand scenarios, which make up very few hours during the year,” said Jim Burke, president and CEO of Vistra. “AI factories that have the flexibility to adjust their power use with grid conditions are a faster solution, especially with co-located generation, for better utilization of the current grid infrastructure. This helps boost speed while we continue to build out more infrastructure for the long term.”
Emerald AI and NVIDIA tested AI power flexibility at five commercial data centers over the past year. DSX Flex is scheduled for commercial deployment in 2026 at the NVIDIA AI Factory Research Center in Virginia, which will use Vera Rubin infrastructure to support power-flexible AI operations.
The companies are advancing projects based on the Vera Rubin DSX reference design with DSX Flex to scale AI infrastructure and support faster grid interconnections. The approach also supports development of new generation capacity, local economic activity, and deployment of AI systems.
Source: NVIDIA
About NVIDIA
NVIDIA, founded in 1993 and headquartered in Santa Clara, CA, designs and manufactures graphics processing units, systems on chips, networking hardware, and AI intelligence software such as CUDA. Its products serve industries including gaming, data centers, autonomous vehicles, professional visualization, robotics, health care, and energy. The company introduced the GPU in 1999 and later expanded into accelerated computing and AI infrastructure. In gaming, its GPUs support high-performance rendering, while in AI and high-performance computing, its systems provide the infrastructure for training and deploying large-scale models. NVIDIA also develops tools for robotics and autonomous driving.
About Emerald AI
Emerald AI is a U.S.-based energy technology company that develops software to manage power use in AI data centers. Founded in 2024, it is headquartered in Washington, DC. Its main product, the Conductor platform, connects data centers with electric grids and adjusts computing workloads in real time. The platform supports AI training and inference while helping maintain grid stability. Emerald AI serves data center operators and utility companies that need to balance energy demand and compute performance. The company operates a B-to-B model and provides software through subscription services. Its technology is used in projects across the United States and Europe, focusing on energy flexibility and system reliability.