OpenAI was an artificial intelligence research lab little known outside of Silicon Valley before ChatGPT debuted in November 2022.

OpenAI in recent weeks has announced partnerships with chipmakers Nvidia and AMD that will supply the AI startup with specialized chips for running its AI systems. OpenAI has also made big deals with Oracle, CoreWeave and other companies developing the data centers where those chips are housed.

A string of massive AI chip investments have highlighted the technology industry’s surging appetite for computing power as it races to build systems that meet or surpass human intelligence.

OpenAI and Broadcom are collaborating to build and deploy 10 gigawatts of custom AI accelerators, scaling global AI infrastructure with enhanced efficiency, performance, and intelligence.

The companies said OpenAI would design the chips, which Broadcom will develop and deploy starting in the second half of 2026. They will roll out 10 gigawatts worth of custom chips, whose power consumption is roughly equivalent to the needs of more than 8 million U.S. households or five times the electricity produced by the Hoover Dam.

OpenAI’s development of its own artificial intelligence chip could end upbenefiting SoftBank, one of its biggest shareholders, which is also helping the ChatGPT maker develop and finance its ambitious data center plan.

Learn why OpenAI’s move into custom AI chip development could redefine the future of artificial intelligence and its applications. Broadcom and

Oracle’s new leadership is standing in front of a $300 billion bet that could reshape the company’s future, or crush it. The dual CEOs, Mike Sicilia and Clay Magouyrk, are defending a massive AI push built on new data centers and an aggressive partnership with OpenAI,