Grid software unicorn Emerald AI has formed a coalition with Google and Nvidia that will deploy the startup’s technology to secure additional grid space for data centers. The initiative is called the AI Energy Management Alliance (AEMA).
What AEMA intends to do
AEMA seeks to make demand response — the practice of temporarily reducing electricity use — a built-in element of data center planning. By pausing noncritical tasks and shifting compute loads, the alliance estimates it could enable up to 100 gigawatts of additional data center capacity to connect to the power grid.
Anthropic has joined the founding trio, and the coalition includes several utilities: AES, Constellation, National Grid and NRG Energy.
Demand response: established idea, new application
Demand response is an established utility practice: grids are designed for peak loads, and for much of the year actual demand is well below maximum capacity. Historically, demand response involved large electricity consumers such as factories agreeing to reduce usage during peak times — for example by pausing production or switching to backup generators — in return for payments from utilities.
Today data centers can participate in these programs, typically by running backup generators. Emerald AI proposes an alternative to diesel generators: its software coordinates utility requests with data centers so they can quickly suspend noncritical workloads or move them to locations with spare grid headroom.
Comparable efforts and the numbers
Emerald AI is not alone. Google has been developing its own tools, and Enel X offers solutions that use data centers’ uninterruptible power supplies (UPS) to smooth power demand peaks. A Goldman Sachs study published last year found that limiting maximum grid usage to 90% for a few hours at a time could free about 76 gigawatts of capacity.
Emerald AI’s advantage is direct software connections between utilities and data centers, enabling rapid responses akin to batteries. The startup recently raised $150 million in a Series A round led by Energize Capital and DCVC, funding that can help scale the technology.
Limits and practical considerations
Given the peaky nature of AI compute loads and data centers’ ability to ramp up and down quickly, AEMA addresses a pressing need. The coalition also aims to help tech companies and utilities identify new data center sites, a longstanding challenge for both sides.
However, more sophisticated demand response will not solve all grid issues immediately. Ayse Coskun, Emerald AI’s chief scientist, told TechCrunch that the company’s technology can reduce the industry’s need for new generation sources, but it will not remove that need entirely.
Overall, AEMA represents a significant step toward more flexible use of existing grid capacity for data centers, while leaving open the broader task of meeting long-term generation and infrastructure needs.
