With a projected capacity of 7 GW, the Stargate data centers will power OpenAI’s next-gen models—fueled by massive investments from Oracle, SoftBank, and Nvidia.
OpenAI’s Stargate Project Just Got a Major Expansion
OpenAI has unveiled plans to build five new AI data centers across the U.S. as part of its Stargate infrastructure project, partnering with Oracle and SoftBank to power its next wave of AI development.
- The new data centers will boost Stargate’s projected capacity to 7 gigawatts (GW) — enough to power over 5 million homes.
- This marks a significant scale-up as OpenAI prepares for larger AI models, increased demand, and real-time inference at global scale.
“The future of AI is as much about compute infrastructure as it is about algorithms,” said one OpenAI executive familiar with the project.
Where the New Stargate Sites Are Located
Three of the five new data centers are being developed in partnership with Oracle:
- Shackelford County, Texas
- Doña Ana County, New Mexico
- An undisclosed location in the Midwest
Two additional data centers are being built with SoftBank:
- Lordstown, Ohio
- Milam County, Texas
Together, these facilities will significantly increase OpenAI’s operational footprint, particularly in energy-rich and rural areas ideal for high-density data infrastructure.
A Massive Power Play: 7 GW and Counting
The 7 gigawatt capacity target makes Stargate one of the largest AI data infrastructure efforts in the world.
- For comparison, 1 GW can power about 750,000 homes.
- This infrastructure will be crucial for training the next generation of multi-modal models, supporting long-context processing, real-time AI applications, and enterprise integrations.
The facilities are expected to host tens of thousands of Nvidia GPUs, with custom-built architectures optimized for energy efficiency and thermal management.
Fueling It All: $100 Billion From Nvidia
This expansion comes just a day after OpenAI announced it would receive a $100 billion investment from Nvidia, aimed at:
- Purchasing cutting-edge AI chips
- Building out additional Stargate infrastructure
- Developing custom silicon solutions for future OpenAI needs
Nvidia’s role is pivotal in enabling OpenAI to keep pace with the explosive computational demand of generative AI, particularly as model sizes and usage scale up dramatically.
Why This Matters
OpenAI’s Stargate project signals a fundamental shift in the AI arms race:
- Compute is the new currency — the companies with the most advanced infrastructure will define the future of AI capabilities.
- By partnering with Oracle and SoftBank, OpenAI is leveraging both cloud scale and global capital to build faster and bigger.
- The regional spread of data centers could also help mitigate latency and improve redundancy across geographies.
This expansion may also trigger regulatory scrutiny over energy usage, land use, and AI compute monopolization.
