A OpenAI is working with Broadcom and TSMC to build its first in-house chip designed to support its systems artificial intelligence (AI) while adding AMD chips alongside Nvidia chips to meet growing infrastructure demands, sources told Reuters.
ADVERTISING
A OpenAI reportedly formed a chip development team of about 20 people, including lead engineers who previously worked on the company's Tensor processors. Google for AI.
Still, on its current timeline, custom hardware may not enter production until 2026.
Meanwhile, sources also said that the OpenAI is incorporating AMD chips into its setup Microsoft Azure. AMD introduced its MI300 chips last year, which was a big part of the news this summer that its data center business doubled in a single year as it chases market leader Nvidia.
ADVERTISING
A The Information had reported in July that the OpenAI was in discussions with Broadcom and other semiconductor designers about developing its own AI chip, and earlier this year, Bloomberg reported that the OpenAI was working to build its own network of foundries, but according to Reuters, those plans have been delayed due to cost and time.
The reported strategy places the OpenAI in a similar path to other tech companies trying to manage costs and access to AI server hardware with custom chip designs. But Google, Microsoft and Amazon are already a few generations ahead in their efforts, and the OpenAI may need a lot more funding to become a true competitor.
Read also
