Emerald AI teams with Google, Nvidia and Anthropic to free up grid capacity for data centers
Emerald AI has formed the AI Energy Management Alliance with Google, Nvidia, Anthropic and four utilities to make demand response standard in data center development, aiming to open 100 gigawatts of new grid capacity.
The group, called the AI Energy Management Alliance, or AEMA, wants to make demand response — a practice in which data centers temporarily cut their electricity use — a standard part of data center development. Pausing noncritical tasks and shifting some computing loads to other locations would allow an additional 100 gigawatts of data centers to connect to the grid, the alliance says.
Anthropic is a founding member alongside Emerald AI, Google and Nvidia. A group of utilities has also joined: AES, Constellation, National Grid and NRG Energy.
Demand response is a decades-old tool for grid operators. It works because grids are built to handle peak loads, while demand stays well below that maximum for most of the year. Factories and other large electricity users have long been paid by utilities to cut consumption when demand spikes, usually by halting production or switching to backup generators. Data centers can qualify for the same programs today, typically by firing up diesel backup generators.
Emerald AI is proposing an alternative to those generators. Its software links utility requests directly to data centers, which can pause noncritical work or move loads to a site where the grid has spare capacity.
The startup is not alone in the field. Google has been building its own tools, and Enel X lets data centers draw on their uninterruptible power supplies to smooth peaks in demand. A Goldman Sachs study published last year found that capping data centers' maximum grid usage at 90 percent for a few hours at a time could free up 76 gigawatts of capacity.
Emerald AI's software connects utilities directly to data centers, which the company says should let facilities respond to requests as quickly as batteries can. The startup recently raised $150 million in a Series A round led by Energize Capital and DCVC, funding it intends to use to roll out the technology more widely.
Given the volatile nature of AI computing loads and the speed at which data centers can ramp power use up and down, TechCrunch AI reported that the creation of a group like AEMA was overdue. The coalition also says it will help technology companies and utilities identify new sites for data centers, an effort that has proved difficult for both sides.
Ayse Coskun, Emerald AI's chief scientist, told TechCrunch that the technology can blunt the industry's need for new generating sources but will not eliminate it entirely.