Emerald AI Raises $150 Million Series A to Make Data Centers Flexible Grid Assets

Washington, D.C.-based Emerald AI closed a $150 million Series A at a $1.05 billion valuation on August 25, co-led by Energize Capital and DCVC with participation from NVIDIA, Samsung Ventures, Siemens, Aramco Ventures, Salesforce Ventures, GE Vernova, RWE, JERA Ventures, In-Q-Tel, Radical Ventures, Energy Impact Partners, and Lowercarbon Capital. The round brings total funding to more than $220 million and is one of the largest Series A financings of 2026. Emerald's product, Emerald Conductor, is software that varies data-center electricity consumption in response to grid conditions while maintaining computing workloads.

How Emerald Conductor Works

Emerald Conductor sits between a data center's workload scheduler and the local utility, modulating compute intensity up or down based on grid frequency, pricing, carbon intensity, and curtailment signals. The company says it has completed five demonstrations and has begun moving toward commercial deployments with hyperscale and colocation customers. The software is intended to convert AI data centers from fixed-load facilities into flexible grid assets that can absorb excess renewable generation, reduce consumption during scarcity events, and earn revenue in capacity markets without sacrificing SLA-bound AI training and inference jobs.

Why Investors Are Paying Up

The investor list — which combines hyperscaler capital (NVIDIA, Salesforce), grid operators (GE Vernova, RWE, JERA), energy incumbents (Aramco, Siemens), and intelligence community capital (In-Q-Tel) — reflects the unusual cross-sector nature of the bet. Emerald's pitch is that data centers are growing fast enough that their electricity demand is now a meaningful share of regional grids, and that flexible load is the cheapest new "peaker" capacity available. The Series A valuation of more than $1 billion for a Series A round places Emerald among the largest infrastructure-software rounds of 2026 and signals investor conviction that the AI-electrification collision is creating a new software category.

Industry Context

Emerald's round is the largest infrastructure-side deal in a busy week for AI-focused startups. Gatik raised $200 million for autonomous freight, Alice raised $140 million for AI security, and Stability AI raised $76 million for generative media, taking the week's ten announced financings to $747.9 million, according to TechStartups. Emerald sits in a category with several competitors including GridBeyond, Voltus, and EnergyHub, but its AI-specific positioning — explicitly targeting LLM training and inference workloads — distinguishes it from older demand-response vendors.

What to Watch Through Year-End

Three checkpoints follow. Emerald's first commercial deployment with a hyperscale or colocation customer, expected to be announced in Q4 2026, will be the first real test of whether Emerald Conductor can deliver on its flexibility promises without disrupting AI training schedules. The launch of utility programs that compensate data centers for flexible load — expected in PJM, ERCOT, and CAISO before year-end — will determine the size of the addressable market. And competition from hyperscaler-internal solutions, particularly NVIDIA's own data-center energy software and the broader push by utilities to offer flexible-tariff products, will reveal whether Emerald remains a standalone category leader or is absorbed into the AI infrastructure stack.