For AI hardware suppliers and investors, Foxconn’s proposed collaboration with OpenAI is a wager that demand can support further server expansion and double-digit growth next year.
Young Liu, Foxconn’s chairman, announced on Friday that the company was developing the new AI hardware collaboration after meeting OpenAI chief executive Sam Altman earlier that day. Liu said Foxconn would make a further announcement once the details were confirmed.
The proposed tie-up would build on an agreement reached in 2023, when Foxconn and OpenAI began working on the design and engineering of data-centre racks, components and other AI hardware.
Foxconn, formerly known as Hon Hai Precision Industry, is the world’s biggest contract electronics manufacturer and Nvidia’s biggest AI server maker. Cloud and networking products, including AI servers, provided 51% of its second-quarter revenue, passing the 50% threshold for the first time, Foxconn said in a statement in August.
Liu said the AI industry was not experiencing a bubble, despite recent stock-market concerns. He expects Foxconn to maintain double-digit growth next year and said the company saw a need to expand its AI server manufacturing further.
That manufacturing base includes six factory buildings in Houston, Texas, where Foxconn makes AI servers for Nvidia. The company has also begun using humanoid robots inside those facilities.
Brand Cheng, chairman of Foxconn Industrial Internet, told Reuters that more than 10 humanoid robots had been deployed at the Texas sites for internal use. The unit introduced two wheeled humanoid models in October: one designed to lift objects vertically, and another able to bend at the waist for assembly, loading and unloading work.
Foxconn Industrial Internet is developing artificial intelligence models for the robots using Nvidia’s GR00T foundation model.
Comments (0)