Emerald AI Series A brings $150M at reported $1.05B valuation
Emerald AI raised $150 million to expand software that lets data centers shift power demand during grid stress.
By Dominic Okoye · Staff Writer
· 3 min read
Emerald AI’s Series A has brought in $150 million at a reported $1.05 billion private-company valuation, giving the data-center power software startup fresh capital to expand commercial deployments. Energize Capital and DCVC co-led the financing, announced Aug. 25, with Nvidia, Samsung Ventures, Siemens, GE Vernova and Salesforce Ventures among the reported participants.
The valuation is the negotiated price attached to this private financing, rather than a public-market measure. It sets the company’s new financing benchmark, though the available reporting does not independently establish whether the figure is pre-money or post-money. For more on how those figures affect ownership and future rounds, see how startup valuation works.
Emerald said it will use the proceeds to scale deployments internationally. The company launched publicly in July 2025, according to SiliconANGLE. Reports differ on cumulative fundraising, with the Washington Business Journal putting it at $215 million since the company’s 2024 founding and SiliconANGLE reporting more than $220 million, so the total cannot be reconciled from the available reporting.
What does Emerald AI’s software do for data centers?
Emerald’s product, Emerald Conductor, is designed to coordinate AI workloads with on-site batteries and power generation. In practical terms, the company says a participating data center can reduce or shift electricity use when a utility or grid operator needs relief, while keeping designated computing jobs running.
That proposition would make a large power user a flexible load, rather than a constant draw on the grid. Utilities could seek verified reductions during stressed periods, while operators could potentially obtain more or faster power access. The model depends on both the software’s ability to manage workloads and agreements that define when a utility can call on that flexibility.
Emerald estimates flexible operation could make more than 100 gigawatts of capacity available on the existing U.S. grid. That is a company estimate, not an independently established outcome in the available material.
Where has Emerald AI deployed its technology?
Emerald has reported five demonstrations on commercial data-center load in Arizona, Illinois, Virginia, Oregon and London, with partners including Nvidia, Oracle and National Grid. SiliconANGLE reported that the company subsequently began a commercial deployment in California.
One concrete utility example is Silicon Valley Power’s Flexible Load Interconnection Program in Santa Clara, California. Under the arrangement described by SiliconANGLE, data centers that commit to verified, dispatchable flexibility can seek faster and larger interconnection. The supplied reporting does not provide results from the program.
Emerald is also working with Digital Realty and Nvidia on the nearly 100-megawatt Vera Rubin AI Research Factory in Manassas, Virginia, which SiliconANGLE said was due online later in 2026. Dominion Energy, PJM Interconnection and the Electric Power Research Institute are helping with tests, according to the report.
The funding is a bet on software-mediated demand management as AI infrastructure runs into power constraints. Emerald has named utilities, AI companies and data-center operators as customers, but the available reporting does not independently establish customer counts, revenue or the economics of its commercial contracts.
This story draws on original reporting from SiliconANGLE.
