OpenAI and Broadcom Announce Custom AI Chip Collaboration
The multi-year agreement will see Broadcom manufacture 10 gigawatts of custom AI accelerators designed by OpenAI, starting in the second half of 2026.
A Strategic Shift to Custom Silicon
OpenAI and Broadcom announced a major strategic collaboration to co-develop and deploy custom artificial intelligence accelerators. The partnership signals a significant move for OpenAI, which has historically relied on GPUs from vendors like NVIDIA. By designing its own chips, OpenAI can optimize hardware for its specific model architectures.
The collaboration positions OpenAI as a fabless chip designer, with Broadcom handling the development and manufacturing of the systems. This follows a separate deal OpenAI announced with AMD for 6 gigawatts of GPUs, indicating an aggressive, multi-pronged strategy to secure massive amounts of computing power.
…Archive Access
This article is older than 24 hours. Create a free account to access our 7-day archive.