Grid software unicorn Emerald AI has formed a coalition with Google and Nvidia to use the startup’s technology to free up space on the grid for more data centers.
The AI Energy Management Alliance (AEMA) wants to make demand response, a paradigm in which data centers temporarily reduce power usage, an integral part of data center development. AEMA says it could potentially connect an additional 100 gigawatts of data centers to the grid by pausing non-essential tasks and shifting some computational loads.
In addition to Anthropic, the founding trio will be joined by a variety of utility companies, including AES, Constellation, National Grid, and NRG Energy.
Demand response is an old concept that has been used by utility companies for decades. This works because the grid is designed with peak loads in mind. For most of the year, demand is well below the maximum that the grid can support.
Traditionally, demand response has been aimed at large electricity users, such as factories, who agree to reduce usage when demand increases, usually by pausing production or switching to backup generators. In return, the utility company pays you, often a significant amount, for their services.
Data centers can now also take advantage of these programs, usually by turning on backup generators.
Emerald AI has proposed an alternative to dirty diesel generators. The company’s software coordinates requests from power companies with data centers, allowing them to pause non-essential tasks or shift load to other data centers that have more power on the grid.
This startup isn’t the only one working on such technology. Google has developed a proprietary tool, Enel X, that allows data centers to take advantage of uninterruptible power supplies to avoid peaks in power demand. A Goldman Sachs study published last year found that data centers could free up 76 gigawatts of capacity by limiting maximum grid usage to 90% for a few hours at a time.
However, Emerald AI also has advantages. The company’s software connects power companies directly to data centers, which should be able to respond to requests as quickly as batteries. The company recently raised $150 million in Series A led by Energize Capital and DCVC, providing the funding needed to deploy the technology more broadly.
Given the peak nature of AI’s computational load and the rapid growth and decline of data centers, the creation of something like AEMA is premature. The coalition also promises to help tech companies and power companies find new sites for data centers, an effort that will not be easy for either party.
Despite its promise, more advanced demand response will not solve all of today’s power grid problems. Ayse Coskun, chief scientist at Emerald AI, told TechCrunch that while the company’s technology promises to slow the industry’s demand for new power generation sources, it won’t eliminate it completely.
If you make a purchase through links in our articles, we may earn a small commission. This does not affect editorial independence.
