Meta signs AI computing deals with Crusoe for 1.6 gigawatts across Texas and Missouri sites

Meta Secures Major AI Computing Capacity from Crusoe

Meta has signed new agreements to obtain AI computing power from data center firm Crusoe, according to Bloomberg News citing people familiar with the matter2

2

. The deals cover two facilities located in Childress, Texas, and Warrenton, Missouri, and will provide Meta with roughly 1.6 gigawatts of capacity combined across both sites3

3

. To put this in perspective, a single gigawatt is enough to power approximately 750,000 U.S. homes4

4

. Neither Meta nor Crusoe immediately responded to Reuters' requests for comment on the report1

1

.

The Shift from Chips to Electricity in AI Infrastructure

The 1.6 gigawatts figure reveals how the unit of AI ambition has fundamentally shifted from chips to electricity. Meta is no longer just buying servers; it is contracting for power at the scale of national grid planning, and doing so across multiple developers simultaneously2

2

. A gigawatt represents the rough output of a large power station, and dedicating 1.6 of them to a single company's computing needs demonstrates the massive infrastructure requirements driving Meta AI expansion. The company intends to use this AI computing capacity for training and serving the AI models and agents it has reorganized itself around2

2

.

Crusoe's Growing Role in AI Infrastructure

Founded in 2018, Crusoe specializes in building and running large data centers for AI work, and adding Meta to its client roster places the company among the developers that the largest AI buyers turn to when they need capacity faster than they can build it themselves2

2

. Crusoe already has deals with Oracle, Microsoft, and Alphabet's Google for three other campuses, making it a counterparty of choice for tech giants racing to secure AI computing power2

2

. The financial terms and timeline for when the computing capacity will be delivered remain unclear, as the people familiar with the matter requested anonymity because the discussions are private4

4

.

Meta's Broader AI Infrastructure Strategy

The Crusoe agreements represent an incremental addition to Meta AI infrastructure buildout. The company's single biggest effort is a nearly 4,000-acre campus in Louisiana designed to provide up to five gigawatts of capacity2

2

. Meta has pledged to invest $600 billion in U.S. infrastructure and jobs over the next three years as it builds out massive data centers to power CEO Mark Zuckerberg's aggressive bets on AI agent technologies3

3

. Like its Big Tech peers, Meta has been pouring billions of dollars into AI data centers as demand continues to outstrip supply3

3

.

Strategic Choice: Speed Over Ownership

The shift toward leasing AI computing capacity from developers like Crusoe, rather than building every site in-house, represents a strategic choice. Constructing gigawatt-scale data centers takes years and ties up enormous capital, while contracting for capacity that someone else builds and operates lets Meta move faster and spread construction risk, at the cost of paying a margin to the developer2

2

. With demand running ahead of supply, speed has become worth more than ownership. Meta's answer has been to spread its bets across developers and geographies rather than rely on any single site, as data centers of this size need power and water on a scale that has begun to draw scrutiny from grid operators and communities2

2

. The competition for electricity is now a defining constraint on the industry, visible in operators racing to secure gigawatts across multiple continents. What these deals confirm is the trajectory: a company whose future depends on Meta's AI ambitions is buying the electricity to power it by the gigawatt, from whoever can deliver it fastest2

2

.