Emerald AI Raises $150M to Make Data Centers Grid-Flexible

Emerald AI, a Washington, D.C.-based startup that helps AI data centers reduce or shift their power consumption when the electrical grid comes under stress, raised $150 million in an oversubscribed Series A round at a $1.05 billion valuation, according to the company's own announcement of the raise. The round was co-led by Energize Capital and DCVC, with participation from Nvidia, Samsung Ventures, GE Vernova, and Salesforce Ventures, bringing Emerald AI's total funding raised to more than $220 million.

Emerald AI's underlying thesis directly challenges how data centers have traditionally operated. Rather than treating data centers as fixed loads consuming roughly the same amount of electricity regardless of grid conditions, the company's software lets AI infrastructure function as a flexible resource that dynamically adjusts power consumption in response to real-time grid conditions, according to Pulse2's reporting on the technology. The product runs on AI agents that autonomously decide how operators and their customers should shift power loads based on grid needs.

Emerald AI's Series A at a Glance

The scale of the opportunity Emerald AI is targeting is substantial. The company estimates that applying flexible-load technology across the AI infrastructure buildout could unlock more than 100 gigawatts of currently untapped capacity on the existing U.S. power grid, according to Pulse2's reporting on the company's projections. That matters given that building new transmission, generation, and grid infrastructure can take many years, while data center electricity consumption is projected to represent a significant portion of U.S. demand growth through 2030. Under Emerald AI's model, data centers can receive expanded, faster access to the grid in exchange for providing verified, dispatchable flexibility whenever the electricity system needs relief.

Emerald AI has moved well past the pilot stage. The company has completed five commercial demonstrations globally and its technology is now deployed commercially at multi-megawatt, full data center scale, serving customers spanning leading AI firms, data center operators, and electric power utilities directly, according to Morningstar's coverage of the announcement. One notable active deployment is in Manassas, Virginia, where Emerald AI is working with Digital Realty and Nvidia on a nearly 100-megawatt Vera Rubin AI Research Factory, built specifically as a power-flexible AI facility.

Emerald AI founder and CEO Dr. Varun Sivaram, who previously held a senior climate position in the Biden administration, framed the company's mission around solving a self-inflicted problem within AI itself. "We founded Emerald AI on the conviction that the intelligence driving the AI revolution could solve its own greatest bottleneck: power," Sivaram said. According to Dealroom's reporting on the raise, Sivaram's stated goal is to make data centers grid-flexible and, in the process, "transform them from villains into heroes." Energize Capital's John Tough offered a blunt summary of the underlying investment thesis: "You can't deliver compute without energy."

Emerald AI's raise arrives at a moment when data center power demand has become a genuine flashpoint, connected directly to the wave of community and municipal backlash we've tracked closely this month, including Essex, Ontario's proposed one-year data centre moratorium and Alberta's contentious town halls over grid and water concerns. If technology like Emerald AI's genuinely reduces the strain new AI data centers place on local grids, that could meaningfully change the terms of the community and regulatory pushback playing out across multiple jurisdictions simultaneously.

Emerald AI's rapid path to unicorn status is worth watching for any business evaluating AI infrastructure investment or partnerships, since grid capacity and power availability have become genuine bottlenecks constraining how quickly new AI compute can come online, a cost and timeline factor connected to Nvidia's own recent 15% AI server price hike driven partly by broader supply chain and infrastructure pressure.

For businesses in utilities, real estate development, or communities weighing data center proposals, Emerald AI's technology is worth understanding as a potential mitigating factor in future negotiations, since flexible-load data centers could plausibly reduce some of the grid strain and cost concerns currently driving local opposition to new facilities.

What does Emerald AI's technology actually do?
Emerald AI's software lets AI data centers dynamically adjust their power consumption based on real-time electrical grid conditions, functioning as a flexible resource rather than a fixed, constant power load.

How much funding has Emerald AI raised?
Emerald AI raised $150 million in a Series A round at a $1.05 billion valuation, bringing its total funding raised since founding in 2024 to more than $220 million.

How much grid capacity could Emerald AI's technology unlock?
The company estimates that applying its flexible-load technology across the AI infrastructure buildout could unlock more than 100 gigawatts of currently untapped capacity on the existing U.S. power grid.

Emerald AI raised $150 million in an oversubscribed Series A round at a $1.05 billion valuation, backed by Nvidia, Samsung Ventures, and Salesforce Ventures among others, to scale technology that lets AI data centers dynamically flex their power consumption based on grid conditions. The company estimates its approach could unlock more than 100 gigawatts of currently untapped capacity on the existing U.S. power grid, with technology already deployed commercially at multi-megawatt scale following five completed demonstrations. The raise lands amid mounting community and regulatory backlash over AI data center power and grid strain, positioning Emerald AI's flexible-load approach as a potential mitigating factor in that ongoing tension.