OpenAI has partnered with Oracle to develop 4.5 gigawatts of additional data center capacity in the United States under the Stargate initiative. This large-scale buildout is said to boost the country’s AI capabilities, create over 100,000 jobs, and help meet OpenAI’s soaring demand for compute.
With this new phase, the total Stargate capacity under development will exceed 5 gigawatts, powered by more than 2 million chips. The new capacity follows OpenAI’s earlier announced plan at the White House to invest $500 billion in 10 gigawatts of AI infrastructure across the U.S. over the next four years. The company is expecting to surpass this target thanks to strong momentum with partners like Oracle and SoftBank.
Along with the AI capabilities, the new AI infrastructure is expected to create jobs across the board from full-time operational roles to indirect jobs in manufacturing and services. A massive chunk of these roles will support the construction workforce, including thousands of electricians and equipment specialists.
According to the official blog post, Stargate I, which was the first facility located in Abilene, Texas, is already partially operational. Oracle started delivering Nvidia’s GB200 AI servers to the site last month. Now OpenAI has started running early training and inference workloads at the facility to test the limits of its next-gen AI models.
Furthermore, OpenAI is also working with SoftBank on additional sites and new data center designs, as part of its long-term strategy to rethink AI infrastructure. Besides, Microsoft will continue to provide cloud services to OpenAI, including through Stargate.
While the U.S. expansion is in full swing, OpenAI is also strengthening international ties. Recently, the company signed an MoU with the UK Government to collaborate on AI adoption and infrastructure. The deal focuses on using advanced AI to improve public services and stimulate economic growth.
For more updates on AI infrastructure and global tech moves, join us on WhatsApp to stay in the loop.
