Google, Nvidia Form Alliance to Expand Data Center Grid Access

A new industry coalition aims to solve one of AI's most pressing infrastructure challenges: finding enough power on the electrical grid to support massive data center expansion.

Emerald AI, a grid software unicorn, has formed the AI Energy Management Alliance (AEMA) with founding members Google, Nvidia, and Anthropic, according to details first reported by TechCrunch. The group also includes major utilities AES, Constellation, National Grid, and NRG Energy.

The alliance's core strategy centers on demand response—a decades-old utility practice that the coalition believes can unlock 100 gigawatts of additional data center capacity on existing grid infrastructure.

How demand response creates grid headroom

Electrical grids are engineered to handle peak demand periods, which means they operate well below maximum capacity most of the time. Demand response programs have traditionally paid large industrial users like factories to reduce consumption during high-demand periods, often by pausing production or switching to backup generators.

Data centers have participated in these programs primarily by firing up diesel generators. Emerald AI offers a different approach: software that coordinates utility requests with data center operations, allowing facilities to pause non-critical computing tasks or shift workloads to other locations where grid capacity is available.

The startup's technology connects utilities directly to data centers, enabling rapid response similar to battery storage systems. A Goldman Sachs study from last year found that limiting maximum grid usage to 90 percent for a few hours at a time could free up 76 gigawatts of capacity.

Why it matters

The AI boom has created an infrastructure bottleneck. Tech companies are struggling to find suitable sites with adequate power availability, while utilities face pressure to support new data centers without building expensive generation capacity. Demand response offers a path to accommodate growth using existing infrastructure—critical as data center power needs continue to surge. The coalition's formation signals that major AI players recognize grid constraints as a fundamental business challenge, not just an operational detail.

Competition and funding

Emerald AI isn't alone in this space. Google has developed its own demand response tools, and Enel X enables data centers to use uninterruptible power supplies to smooth power demand peaks. However, Emerald AI recently secured $150 million in Series A funding led by Energize Capital and DCVC, providing resources to scale its technology across the industry.

The coalition also plans to help members identify viable sites for new data centers, addressing a challenge that has complicated expansion plans for both technology companies and utilities.

Limitations acknowledged

Ayse Coskun, Emerald AI's chief scientist, told TechCrunch that while the technology can reduce the need for new power generation, it won't eliminate that requirement entirely. The peaky nature of AI computing workloads and data centers' ability to quickly adjust operations make them well-suited for sophisticated demand response programs, but additional generating capacity will still be necessary to support industry growth.

These details were first reported by TechCrunch.