Broadcom and Meta Extend Partnership to Jointly Develop AI Accelerator Chips

On April 14 local time, Broadcom officially announced that it has extended its partnership with Meta through 2029, and will support Meta's initial AI deployment of over one gigawatt of computing capacity. 

Broadcom and Meta plan to launch the industry's first 2nm AI compute accelerator. Over the next three years, the companies are collaborating on next generations of AI accelerator chips to meet Meta's escalating compute demands for next-generation AI models.

Meta Training and Inference Accelerator(MTIA) is a key pillar of Meta's broader silicon strategy, which deploys different accelerators for different workloads — the MTIA portfolio matches purpose-built hardware to optimize both performance and total cost of ownership at scale.

As Meta advances the deployment of its AI data centers, Broadcom will provide XPUs and related technologies to power the MTIA chips. Broadcom will also provide Ethernet networking solutions to help Meta increase computing bandwidth within a single MTIA server rack. 

Given the scale of this expanded partnership, Broadcom President and CEO Hock Tan will step down from Meta's Board of Directors and will serve as an advisor to Meta, providing guidance on Meta's custom silicon roadmap and future infrastructure investments.