OpenAI aims to cut hardware costs by up to 30% through its partnership with Broadcom
OpenAI expects to cut hardware costs by 20 to 30 percent through a joint chip development program with Broadcom, according to a Bloomberg report citing a person familiar with the company’s plans. The custom chips are scheduled to roll out by late 2026 and are part of a multibillion-dollar project valued at several tens of billions of dollars.
Typically, OpenAI budgets around $50 billion to build a 1-gigawatt data center, with roughly $35 billion spent on advanced chips. The new partnership aims to reduce those chip costs significantly.
But the plan comes with major risks. Developing custom silicon requires billions in investment, specialized technical expertise, and multiple design cycles. The rapid pace of AI innovation also increases the risk that the chips could become obsolete quickly. "There’s a steep learning curve," said semiconductor analyst Cody Acree from Benchmark.
AI News Without the Hype – Curated by Humans
Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.
Subscribe nowRead on for the full picture.
Subscribe for hype-free coverage.
- Full access to every article on THE DECODER
- No ads
- Join the comments and community discussions
- A weekly AI news recap via mail
- 6x/year: "AI Radar" — deep dives on the AI topics that matter most
- Daily AI news, always up to date
- Our full ten-year archive
- Covered by a team with 10+ years in AI