Overview

• Rollout of the custom AI accelerator and network systems will begin in the second half of 2026 with completion targeted by the end of 2029.
• OpenAI will design the accelerators and systems, while Broadcom will develop and deploy Ethernet-based racks across OpenAI facilities and partner data centers.
• The companies said they have worked together on the chip and system designs for about 18 months, and they did not disclose financial terms.
• Broadcom shares jumped more than 10% after the announcement.
• The pact expands OpenAI’s multi-vendor compute commitments to roughly 33 gigawatts, a scale Reuters equated to the electricity needs of more than 8 million U.S. households.