OpenAI has officially confirmed it signed a $30 billion per year agreement with Oracle to lease 4.5 gigawatts of new data center capacity in the U.S. This forms part of the multi-billion-dollar Stargate AI infrastructure project announced earlier this year in partnership with Oracle and SoftBank.
🔍 Deal Overview
- Size & Scope
The annual $30 billion commitment is tied to a 4.5 GW buildout—equivalent to the power output of two Hoover Dams—designed explicitly for Stargate, OpenAI’s $500 billion infrastructure plan - Strategic Move
Oracle’s SEC filing in late June disclosed the deal but didn’t name the customer. OpenAI later confirmed it’s the party involved, embracing vendor diversification beyond Microsoft Azure - Building Scale & Timeline
Capacity is being built across multiple U.S. states—including Texas (Abilene), Michigan, and Georgia—as well as expansion of the existing Abilene site. Oracle expects to invest about $25 billion in capex next year to support this growth
🧭 Why It Matters
- Record‑Breaking Cloud Deal
At $30 billion per annum, this is one of the largest cloud infrastructure agreements ever and exceeds Oracle’s total cloud revenue in fiscal 2025 ($24.5 billion) - Infrastructure Megaproject
This forms half of Stargate’s initial capacity for the U.S. build-out and represents a game-changing industrial-scale AI data center ecosystem Financial Times - Market & Stock Impact
Oracle stock hit record highs following the news, with analysts projecting 50%+ cloud infrastructure growth by fiscal 2028 and boosting revenue forecasts past $100 billion by 2029 - Diversifying Compute Supply
This partnership illustrates OpenAI distancing itself from exclusive dependence on Microsoft Azure by adding multi-provider redundancy (Oracle, Google Cloud, CoreWeave) to support its scale
🌐 Global & Strategic Impacts
- U.S. AI Leadership & Jobs
Stargate supports U.S. Reindustrialization efforts with projected job creation—over 100,000 roles in construction and operations as part of the 4.5 GW expansio - Oracle’s Strategic Pivot
Oracle is rapidly scaling AI cloud infrastructure—investing in chips and data centers to compete against Amazon, Microsoft, and Google in high-performance AI workloads - Semiconductor Surge
The collaboration entails commitment to $40 billion worth of Nvidia GB200 chips, solidifying Nvidia’s dominance and feeding OpenAI’s compute-intensive needs
