Nvidia, Google and Emerald AI launch alliance to manage data-center power demand

  • Nvidia, Google and Emerald AI launched the AI Energy Management Alliance.
  • Lambda increased HGX B200 cluster inference throughput 24% within a fixed power budget.
  • Emerald AI reduced Santa Clara facility demand from 4 megawatts to 3 megawatts automatically.

Nvidia, Google and Emerald AI announced the AI Energy Management Alliance on September 16, 2026, aiming to establish technology-neutral standards for data centers that can adjust power consumption in response to grid conditions. The initiative comes as Nvidia promotes DSX MaxLPS, a power-management system validated by Lambda on a 19-node Nvidia HGX B200 cluster that delivered roughly 5 million inference tokens per second within the power budget normally used by 16 full-power nodes. Nvidia said the test increased cluster throughput by 24% and performance per watt by 23%. Emerald AI has also demonstrated automated demand response at its Santa Clara, California facility, reducing demand from 4 megawatts to 3 megawatts in response to more than 200 Silicon Valley Power signals. The alliance says flexible operations can support faster interconnection while helping utilities manage periods of grid stress as global data-center electricity use rises.

The information on this website is generated using AI and we cannot guarantee its accuracy. Please use it as reference information only.