Greg Brockman says OpenAI’s tech found chip optimizations that would’ve taken humans weeks

Greg Brockman says OpenAI used its models to design chips with Broadcom, finding optimizations that human designers would’ve needed weeks to spot.

Steve Jennings/Getty Images for TechCrunch

OpenAI used its AI models to help design custom chips with Broadcom, said Greg Brockman.
The OpenAI president said the AI found chip optimizations that would’ve taken humans weeks to spot.
OpenAI has been building that expertise in-house to better understand the chip-design process, he added.

OpenAI’s models did what chip engineers could do — only much faster, said OpenAI president Greg Brockman.

Brockman said on an episode of “The OpenAI Podcast” published Monday that the company’s models helped uncover chip design optimizations that would have taken human designers weeks to identify.

“We’ve been able to apply our own models to designing this chip,” Brockman said, referring to OpenAI’s new custom silicon developed in partnership with Broadcom.

“You take components that humans have already optimized and just pour compute into it, and the model comes up with its own optimizations,” he added.

Brockman said the AI-assisted process led to “massive area reductions” on the chip — meaning smaller, more efficient hardware — and shaved weeks off the production schedule.

“I don’t think any of the optimizations that we have are ones that human designers couldn’t have come up with,” he said.

“Our experts take a look at it later, and say, ‘Yeah, this was on my list,’ but it was like, 20 things that would’ve taken them another month to get to,” he added.

Brockman said OpenAI has been building that expertise in-house to better understand the chip-design process.

His comments came as OpenAI announced a new partnership with Broadcom to co-design custom silicon and build out AI infrastructure and computing power.

The companies said on Monday that they plan to roll out 10 gigawatts worth of custom chips, which Broadcom would deploy by the second half of 2026 and finish in 2029. They will be spread across OpenAI’s facilities and partner data centers.

“Developing our own accelerators adds to the broader ecosystem of partners all building the capacity required to push the frontier of AI to provide benefits to all humanity,” OpenAI CEO Sam Altman said in Monday’s statement.

Brockman and OpenAI did not respond to a request for comment from Business Insider.

Read the original article on Business Insider

Content Accuracy: Keewee.News provides news, lifestyle, and cultural content for informational purposes only. Some content is generated or assisted by AI and may contain inaccuracies, errors, or omissions. Readers are responsible for verifying the information. Third-Party Content: We aggregate articles, images, and videos from external sources. All rights to third-party content remain with their respective owners. Keewee.News does not claim ownership or responsibility for third-party materials. Affiliate Advertising: Some content may include affiliate links or sponsored placements. We may earn commissions from purchases made through these links, but we do not guarantee product claims. Age Restrictions: Our content is intended for viewers 21 years and older where applicable. Viewer discretion is advised. Limitation of Liability: By using Keewee.News, you agree that we are not liable for any losses, damages, or claims arising from the content, including AI-generated or third-party material. DMCA & Copyright: If you believe your copyrighted work has been used without permission, contact us at dcma@keewee.news. No Mass Arbitration: Users agree that any disputes will not involve mass or class arbitration; all claims must be individual.