FundraiseSep 17, 2026, 10:50 UTC

Crusoe raises $3.9B to build truck-ready data centers for AI

What's the deal? Crusoe, the startup that built one of the world's largest AI supercomputers, has raised $3.9 billion in a Series F round that values it at about $30.9 billion. The round was co-led by Atreides Management, Valor Equity Partners, and Mubadala Investment, with Founders Fund, TPG, and the Qatar Investment Authority also taking part.

Why now? Crusoe spent two years building a data-center complex in Abilene, Texas, used by OpenAI and Microsoft. Now it is going small: manufacturing compact data centers, called Spark, in its own factories, loading them onto flatbed trucks, and deploying them wherever it can find power.

What's the endgame? The pivot reflects a shift in how AI firms access computing. Giant clusters excel at training models, but serving them to users — inference — can run on far fewer chips. "You don't actually need an Abilene to do that," said chief executive Chase Lochmiller.

The Spark units also offer speed. Rather than months of construction and coordinating with local communities, Crusoe ships prebuilt facilities to sites with spare power — a growing edge as labor and electricity become the biggest constraints on development.

How does Crusoe make money? Unlike rival neoclouds that mostly buy chips and rent access, Crusoe earns at several layers: developing and leasing data centers, renting the GPUs inside them, and running AI models for customers. "We sell data centers, GPUs and tokens," Lochmiller said.

Founded in 2018, Crusoe first placed computing equipment near oil-and-gas sites to tap excess energy, serving cryptocurrency miners before the AI boom redirected it toward Nvidia GPUs. The Abilene campus, with roughly 2.1 gigawatts of planned capacity, made it a central player in AI infrastructure. Cloud customers now include Perplexity, and its managed inference business is starting to reach meaningful scale.

The signal: The size of this raise stands out — among all-time Series F rounds for US energy companies, it sits in the 99th percentile. It signals investor conviction that the next infrastructure race is about deploying compute fast, wherever power is available.