OpenAI and Broadcom Partner to Deploy 10 Gigawatts of Custom AI Chips by 2029
Summary
OpenAI and Broadcom announce a landmark multi-year partnership to co-develop and deploy 10 gigawatts of custom AI accelerator chips by 2029, with rack deployments beginning in 2026 to meet surging global AI demand for OpenAI's 800 million weekly active users.
Key Points
- OpenAI and Broadcom announce a multi-year strategic collaboration to co-develop and deploy 10 gigawatts of custom AI accelerators, with rack deployments set to begin in the second half of 2026 and complete by end of 2029.
- OpenAI is designing its own AI accelerator chips to embed insights from frontier model development directly into hardware, while Broadcom provides end-to-end Ethernet, PCIe, and optical connectivity solutions for scale-up and scale-out networking.
- The partnership aims to meet surging global AI demand across OpenAI's facilities and partner data centers, supporting OpenAI's mission to advance artificial general intelligence for its over 800 million weekly active users.