NVIDIA, Google and Emerald AI form coalition to speed AI data center grid connections

NVIDIA, Google and Emerald AI have launched the AI Energy Management Alliance (AEMA) to push for faster electricity grid connections for AI data centers. The coalition ties quicker access to requirements for more flexible power use during peak demand and grid disruptions.

AEMA’s goal: “controllable” AI load

The alliance is aimed at making data centers less disruptive to the electric grid. NVIDIA says US power infrastructure was built for “flat, static electricity demand,” not for computing centers that can adjust consumption.

The coalition wants a standard for facilities that can shift workloads, discharge storage and respond to system contingencies, turning them into a “controllable resource rather than an inflexible load.”

Faster connections in exchange for flexibility

AEMA’s model is reciprocal. In return for faster utility connections, data centers would reduce power use during peak hours and follow operating rules during brief grid disturbances.

Emerald AI CEO Varun Sivaram wrote that new US data centers can wait a decade or more for a grid connection because utilities must ensure enough capacity for the highest-demand periods, including “sweltering afternoons when air conditioners are running full tilt.” He added that the grid is “only 50 percent utilized on average,” and argued that flexible AI data centers could unlock “100GW on our existing grid for flexible data centers.”

Proposed principles and standardization

AEMA is suggesting principles such as staying connected during short grid events, reducing power use when needed and responding to emergencies. The coalition also calls for standard technical requirements, performance metrics and operational data sharing.

The article cites technologies that could support those goals, including batteries to carry facilities through peak periods, on-site generation to support demand and software that can “slow, pause or shift less urgent computing.”

Utility operator incentives and policy push

The coalition also raises the possibility of cost sharing between data centers and grid operators, depending on how flexibility is implemented.

Sivaram said AEMA will press the case “in state capitals and in Washington,” asking governors, regulators and federal policymakers to offer faster and larger grid connections for data centers that commit to flexibility and are held to it.

Why it matters

AI data center growth is running into grid connection delays, and AEMA is seeking to resolve that bottleneck by making flexibility a condition for faster interconnection. The alliance frames its approach as grid-friendly operations that can expand capacity without major new infrastructure timelines.