Emerald AI, a software company aiming to enhance grid flexibility for AI-powered data centers, has successfully closed a $25 million funding round. The investment was led by Nvidia's NVentures and included contributions from Eaton, GE Vernova, Salesforce, Siemens, and IQT, the venture capital arm of the CIA. This latest round brings Emerald AI's total funding to $68 million since its inception 16 months ago.
Addressing Grid Congestion
Founded by Varun Sivaram, Emerald AI seeks to solve the growing challenge of connecting energy-intensive AI factories to the U.S. power grid. As AI development accelerates, data centers are consuming an increasing share of the nation's power supply—a figure projected to rise from less than 5% to 25% within a decade. Sivaram's solution leverages software to dynamically manage power consumption during peak demand periods, ensuring reliability while reducing strain on the grid.
'We don’t have a supply problem; we have a peak problem,' said Joe Dominguez, CEO of Constellation Energy, one of Emerald's partners.
Partnerships and Pilot Programs
Emerald AI has partnered with leading U.S. power producers, including AES, Constellation Energy, Invenergy, NextEra Energy, and Vistra, to test its grid optimization software. Later this year, the company plans to open its first power-flexible commercial AI factory in Virginia, a 96-megawatt facility developed in collaboration with Nvidia.
The initiative aims to unlock up to 100 gigawatts of additional grid capacity, equivalent to powering 75 million homes, by optimizing existing infrastructure. Marc Spieler, Nvidia's senior managing director for global energy, highlighted the potential of Emerald's approach: 'Our goal is to have as much connected to the grid as possible and not go behind the meter, not being islanded, by being flexible.'
Emerald AI's funding and partnerships underscore the critical role of grid optimization in sustaining America's AI infrastructure while prioritizing domestic energy efficiency and reliability.