엔비디아, 구글, 에메랄드 AI는 데이터센터가 전력 사용을 동적으로 관리하고 그리드 운영과 통합할 수 있도록 하는 데 중점을 둔 AI 에너지 관리 얼라이언스(AEMA)를 출범했다.

On Wednesday, Emerald AI, Google, and NVIDIA announced the formation of the AI Energy Management Alliance (AEMA) to advance flexible AI data centers capable of adjusting their power consumption in response to grid conditions.

The coalition stated that greater power flexibility could help AI facilities connect to existing infrastructure more quickly while reducing pressure on the grid and supporting energy affordability. According to AEMA, conventional interconnection processes were designed for facilities with relatively stable electricity demand, making them less suited to AI data centers that can actively manage their power use. Flexible facilities can shift computing workloads, deploy storage, utilize paired generation, and respond to system emergencies, allowing them to function as controllable electricity resources rather than fixed loads.

AEMA plans to establish technology-neutral standards based on the actual grid services a facility can provide. The proposed framework will address response speed, duration, predictability, and emergency performance, alongside requirements for staying connected during short disruptions, reducing consumption when requested, and responding to contingencies. The alliance also plans to standardize technical data and create risk-adjusted interconnection pathways for facilities with verifiable flexibility.

The coalition will bring together companies and organizations spanning the AI and energy industries, including data center operators, technology providers, power producers, utilities, and regional grid operators. Members will work with utilities on interconnection approaches and develop technical, operational, and policy frameworks for grid-responsive AI demand.

AEMA noted that Emerald AI and NVIDIA are already developing AI factories with energy and infrastructure partners that can respond to grid conditions in real time. The new alliance aims to broaden those efforts across the US as policymakers and industry participants develop rules for meeting the growing power needs of AI infrastructure.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy .