OpenAI has secured a new hardware partner. The AI research lab announced on Monday that it formed a partnership with the semiconductor company Broadcom for ten gigawatts worth of custom AI accelerator hardware. These AI accelerator racks will be deployed to OpenAI data centers and partner data centers starting in 2026 and running through 2029.
The company stated in a press release that by designing its own chips and systems, OpenAI can embed what it has learned from developing frontier models and products directly into the hardware, unlocking new levels of capability and intelligence. While the terms of the deal were not disclosed, the Financial Times estimated it could cost OpenAI between 350 billion and 500 billion dollars.
This partnership is the latest in a series of infrastructure deals for OpenAI in recent weeks. Last week, OpenAI announced it was purchasing an additional six gigawatts of chips from AMD in a deal worth tens of billions of dollars. In September, Nvidia announced it was investing 100 billion dollars into OpenAI alongside a letter of intent for OpenAI to use ten gigawatts worth of Nvidia hardware.
OpenAI also allegedly signed a historic 300 billion dollar cloud infrastructure deal with Oracle in September, though neither company has confirmed that arrangement.