Microsoft To Use OpenAI’s Custom Chip Work To Help In-House Effort

Microsoft Corp. plans to use access to OpenAI’s development of custom AI semiconductors to help its own chip effort.

“As they innovate even at the system level, we get access to all of it,” Microsoft Chief Executive Officer Satya Nadella said on a podcast appearance when asked about the development of in-house chips. “We first want to instantiate what they build for them, but then we’ll extend it.”

Under a revised agreement between the two companies, Microsoft has access to the ChatGPT maker’s models through 2032 and research through 2030, or until a panel of experts agrees that artificial general intelligence has been achieved. Microsoft’s intellectual property rights exclude OpenAI consumer hardware.

OpenAI plans to custom-design chips and networking hardware with Broadcom Inc. Microsoft has worked to develop in-house chips, but has seen less success in the effort than cloud rivals such as Alphabet Inc.’s Google.

Microsoft will work with designs from OpenAI and its own team “knowing we have the IP rights,” Nadella said on the podcast released Wednesday hosted by author Dwarkesh Patel.