OpenAI Partners with Broadcom to Build Custom AI Chips in Major Multi-Year Deal
OpenAI has announced a major partnership with Broadcom to create its first custom artificial intelligence (AI) processors. This agreement follows earlier deals OpenAI made with chipmakers Nvidia and AMD, showing its commitment to securing all the computing power it needs for its growing AI services.
Under the new multi-year deal, OpenAI and Broadcom will work together to develop custom AI accelerators and networking systems for next-generation data centers. They plan to deploy 10 gigawatts' worth of these powerful chips starting in the second half of 2026 and continuing through the end of 2029.
The collaboration is expected to involve multibillion-dollar investments and will include Broadcom delivering racks of AI chips and Ethernet networking gear, helping OpenAI scale up its computing clusters for advanced AI models.
OpenAI needs more powerful and efficient chips to support its AI services. By working with Broadcom, OpenAI can design custom processors that better meet its needs.
Broadcom will benefit from a large, multi-year order for its custom chips and networking products, which helps drive its business and stock value.
OpenAI and Broadcom plan to start deploying the new chips in the second half of 2026, with the rollout continuing through 2029.
No. OpenAI has agreements with Nvidia and AMD and this deal adds to those partnerships. Custom Broadcom chips will enhance OpenAI's options for meeting its growing computing needs.
The Motley Fool
Market Watch
CNBC
Reuters
NYTimes
GlobeNewsWire
WSJ
CNBC Television
Barrons
Proactive Investors
Invezz
Investors Business Daily
Investopedia
Seeking Alpha
Benzinga
Yahoo Finance
Schaeffers Research
Bloomberg Technology
Bloomberg Markets and Finance
CNET
24/7 Wall Street
Zacks Investment Research
Forbes
Business Insider
TechXplore