AI infrastructure startup Crusoe has announced plans to build a massive 900-megawatt data centre campus in Texas to support Microsoft’s growing artificial intelligence workloads. The project highlights the accelerating race among technology giants to expand computing capacity for next-generation AI systems.
The new facility will be in Abilene, Texas, adjacent to Crusoe’s existing AI infrastructure. Once completed, the addition will significantly expand the site’s total projected capacity to 2.1 gigawatts, positioning it among the largest AI data centre hubs in the United States. On average, one gigawatt of power is sufficient to supply electricity to approximately 750,000 homes, underscoring the immense scale of the project.
The development comes as part of Microsoft’s broader push to scale its AI capabilities, with the company investing heavily in infrastructure to support services such as Copilot and other generative AI applications. Microsoft reported $37.5 billion in capital expenditure for the quarter ending December 31, with nearly two-thirds allocated to computing chips and AI-related infrastructure.
Crusoe secured Microsoft as a tenant after earlier negotiations with Oracle and OpenAI for the same site did not materialize. Despite the shift, existing partnerships between Oracle and OpenAI in nearby facilities remain intact, reflecting the dynamic and rapidly evolving nature of AI infrastructure deals.
The project will feature an integrated on-site power plant designed to enhance grid reliability and ensure uninterrupted operations for energy-intensive AI workloads. Additionally, the facility will incorporate a “closed loop” cooling system that recirculates water, addressing the high thermal demands of advanced computing systems while improving efficiency.
Construction is already underway, with the first building expected to become operational by mid-2027. The initiative underscores how locations like Abilene are emerging as key hubs for AI development, driven by access to power, land, and infrastructure capable of supporting large-scale data centres.
The investment reflects a broader industry trend where hyperscale’s are committing billions of dollars to build AI infrastructure at unprecedented scale. As demand for training and deploying large language models continues to surge, projects like this are expected to play a critical role in shaping the future of artificial intelligence and global computing capacity.
