The Stargate project is the United States’ first data center devoted to training and developing advanced AI models. The Abilene, Texas, facility will house up to 400,000 Nvidia AI chips. This new AI infrastructure is a venture spearheaded by OpenAI, Oracle, and SoftBank. Mid-2026 marks the expected completion date. The joint venture announcement occurred during a January White House event with President Donald Trump.
Construction Progress and Local Impact
Though OpenAI publicly announced Stargate in January, the groundwork for the Abilene data center began well before that.
“If you came here a year ago, this would be a field of mesquite trees and shrubs,” said Crusoe CEO Chase Lochmiller. “We broke ground in June of last year and have been on a very accelerated build pace.” Crusoe is the site developer.
Michael McNamara, CEO of energy startup Lancium, first secured agreements to build a data center in Abilene in 2021. Around 2,000 workers are currently involved in the site’s construction, which could rise to nearly 5,000. The completed complex will feature eight data center buildings with a capacity to house up to 50,000 Nvidia GB200 semiconductors.
Billions Invested in AI Infrastructure
By the end of 2026, OpenAI and Oracle plan to install 64,000 Nvidia GB200 semiconductors at the Abilene site, with an initial phase of 16,000 chips set for completion this summer. Nvidia has not disclosed the pricing of the GB200 model, but CEO Jensen Huang previously stated that its less powerful B200 chip costs between $30,000 and 40,000 each. That suggests a staggering investment worth billions of dollars for the first phase of Stargate.
The Stakes for AI Development
Stargate is part of a broader effort to address one of AI’s biggest bottlenecks: computing power. As AI models grow more complex, they require large data centers with immense energy capacity to support training and deployment.
This push for AI supercomputing comes amid broader national discussions about AI’s role in global competition. OpenAI CEO Sam Altman has emphasized the need for the US to lead in AI development.
OpenAI’s partnership with Microsoft previously limited the use of Microsoft’s cloud computing services. However, Stargate signals a shift. The venture allows OpenAI to diversify its infrastructure, which reduces reliance on a single provider. Microsoft remains a key investor in OpenAI but now shares its AI development resources with other cloud partners.
What’s Next for Stargate?
Construction is already underway, and funding is being deployed gradually. Of the $500 billion announced for the Stargate initiative, OpenAI has stated that $100 billion covers the initial phase, with further investments rolling out over the next four years.
Not everyone is convinced of the project’s financial backing. “They don’t actually have the money,” Elon Musk commented on X. He suggests that SoftBank has secured less than $10 billion.
Additionally, the competition to secure Nvidia’s latest AI chips has intensified. For instance, Musk’s xAI recently signed a $5 billion deal with Dell Technologies to build AI servers for a supercomputer in Memphis.
OpenAI, however, remains confident the project will progress as planned.
Moreover, OpenAI and SoftBank have scouted additional sites in Pennsylvania, Wisconsin, Oregon, and Salt Lake City for future data centers. With Stargate, OpenAI and its partners aim to establish a network of AI supercomputing hubs that could define the next generation of artificial intelligence.
