NVIDIA, Google and Emerald AI Form Alliance to Speed Up Data Center Grid Connections

NVIDIA, Google and Emerald AI have created the AI Energy Management Alliance (AEMA), a coalition pushing for faster grid connections for AI data centers in return for making those facilities more flexible in their power use.

The group argues that US power infrastructure was designed for steady, unchanging demand, not for computing centers that can adjust how much electricity they draw. Under the proposal, data centers would shift workloads, discharge stored energy and react to grid emergencies, turning themselves into a controllable resource instead of an inflexible load.

Emerald AI CEO Varun Sivaram noted that new US data centers can wait a decade or more for a grid connection, while the grid runs at only about 50 percent utilization on average. If AI data centers were flexible during the grid's worst hours, he wrote, America could unlock 100GW on the existing grid.

AEMA's principles include clear rules for staying connected during brief disturbances, cutting power use when needed and responding to emergencies, along with standardized technical requirements and data sharing. In exchange, grid operators would connect data centers faster and potentially share costs. Proposed technologies include batteries, on-site generation and software that can slow, pause or shift less urgent computing.

The coalition plans to lobby state capitals and Washington, asking governors, regulators and federal policymakers to offer faster, larger grid connections to data centers that commit to flexibility. AEMA does not address the pollution, noise or community disruption associated with data centers, and given public anger over them, the pitch may not land with everyone.

Source: Engadget