OpenAI, SoftBank, Oracle lead $500B Project Stargate to ramp up AI infra in the US

Several large technology firms including OpenAI, SoftBank, Oracle, Nvidia, and MGX have partnered to set up a new company in the US to ramp up AI infrastructure in the country.

Named Project Stargate, the new company will see an investment of $500 billion over the next four years to support the US’ quest to achieve artificial general intelligence (AGI) — a phase where a machine can think and act like a human being — faster than competing nations, such as China.

The project, which will be financially led by SoftBank and operationally by OpenAI, will see an immediate investment of $100 billion. Masayoshi Son will be the chairman of the company.

The company will have Arm, Microsoft, Nvidia, Oracle, and OpenAI acting as the key technology partners, the partners said in a joint statement, adding that currently data centers were being built in Texas for the same.

During a White House press briefing declaring Stargate, Oracle Executive Chairman Larry Ellison said that at least 10 data centers were currently being built, each building having an area of at least half a million square feet.

These companies are also looking to evaluate potential sites across the country for more campuses and Ellison added that there were plans to build 20 more data centers.

Microsoft has first rights of refusal for AI infrastructure

The development also means that Microsoft will no longer be the only exclusive partner for OpenAI when it comes to infrastructure for training AI or related workloads.

However, in a blog post published this week, Microsoft has clarified that it will have the right of first refusal.

“This new agreement also includes changes to the exclusivity on new capacity, moving to a model where Microsoft has a right of first refusal (ROFR). To further support OpenAI, Microsoft has approved OpenAI’s ability to build additional capacity, primarily for research and training of models,” the company wrote.

In fact, the other elements of the partnership, slated till 2030, between Microsoft and OpenAI are also not changing.

Microsoft will have the rights to OpenAI IP, inclusive of both model and infrastructure, for use within its products, such as the Copilot, the company wrote, adding that the OpenAI API is exclusive to Azure, will run on Azure, and is also available through the Azure OpenAI Service.

Moreover, both companies will continue to honor their revenue-sharing agreements, flowing both ways.

As part of Stargate, Oracle and Nvidia will collaborate with OpenAI on building the computing system for the project, the partners said.

Trump to help with energy requirements

Other than racing towards AGI, Stargate’s infrastructure is expected to create “hundreds of thousands of American jobs” and support the re-industrialization of the US while making it capable of protecting its national security.

During a White House press briefing, US President Donald Trump said Stargate was designed to keep the US ahead of countries such as China.

Earlier in the day, the President repealed a 2023 executive order by Joe Biden that mandated stricter oversight of artificial intelligence (AI) development.

The rationale behind the repeal was that Biden’s executive order though focused on guarding against potential AI risks, placed undue burdens on companies and stifled technological advancement.

During the briefing on Stargate, President Trump also said that he would help Stargate’s partner companies meet the electricity and energy requirements to build out the data centers.

Trump’s promise holds significance as data center power constraints and burgeoning AI workloads have companies scrambling to find new sources of electricity. An approximate 40% of existing AI data centers will be “operationally constrained by power availability” by 2027, a report by market research firm Gartner showed. The firm estimates that the demand from incremental AI servers will reach 500 terawatt-hours per year by then, which is 2.6 times the 2023 level.