Microsoft has agreed to lease a large-scale data center project in Abilene, Texas, originally intended for Oracle and OpenAI, according to Bloomberg report. The facility is designed to deliver approximately 700 megawatts of capacity, underscoring the scale of infrastructure required for modern AI workloads.
The site is located near Oracle and OpenAI’s Stargate campus, a flagship hub for AI computing. Microsoft reached an agreement with developer Crusoe after earlier discussions with Oracle and OpenAI did not result in occupancy commitments. However, sources indicate that OpenAI’s broader partnership agreements with Oracle remain unchanged.
The development reflects shifting priorities among AI companies as they adjust capacity planning, financing strategies, and long-term infrastructure needs. Earlier reports suggested that Oracle and OpenAI had reconsidered expansion plans at the Abilene site due to prolonged negotiations and evolving requirements.
Demand for data center capacity continues to accelerate as companies deploy generative AI systems such as ChatGPT and Microsoft’s Copilot. These services require significant compute power, driving sustained investment in hyperscale infrastructure, including energy, cooling, and semiconductor supply chains.
Microsoft has been expanding its global data center footprint to support AI services integrated across its cloud and enterprise platforms. The Abilene project adds to a growing pipeline of facilities aimed at securing long-term compute capacity.
The broader industry trend highlights intensifying competition for AI infrastructure. Tesla and SpaceX recently announced plans for “Terafab,” a joint AI chip factory initiative, signaling parallel investments across both compute hardware and data center ecosystems.
