In a strategic move set to fundamentally reshape the landscape of artificial intelligence computing, OpenAI and Broadcom Inc. have announced a monumental collaboration. The partnership is centered on the co-development and deployment of a staggering 10 gigawatts of custom AI accelerators, with mass-scale deployment of complete system racks targeted to commence in the second half of 2026 and conclude by the end of 2029. This initiative represents one of the most significant private infrastructure projects in the history of the AI industry, aimed directly at meeting the explosive global demand for advanced computational power.
This deep collaboration moves beyond a simple supplier relationship into a true co-development venture. OpenAI will take the lead on designing the bespoke AI accelerators and overarching systems, directly embedding the profound lessons learned from building its frontier models like GPT-4 and ChatGPT directly into the hardware’s architecture. Broadcom will then bring these designs to life, developing the chips and integrating them with its industry-leading networking solutions. This end-to-end approach allows OpenAI to optimize the entire stack for its specific AI workloads, a critical advantage in the race toward more capable and efficient models.
The physical backbone of this new infrastructure will be massive racks of AI systems, all scaled using Broadcom’s robust portfolio of Ethernet-based networking and connectivity solutions. This strategic choice positions open, standards-based Ethernet as the definitive networking technology for scaling AI data centers, both for scale-up within a cluster and scale-out across multiple facilities. The deployment will span OpenAI’s own data centers as well as those of its partners, creating a distributed and powerful network for next-generation AI processing.
Leadership from both companies emphasized the transformative nature of this partnership. “Partnering with Broadcom is a critical step in building the infrastructure needed to unlock AI’s potential and deliver real benefits for people and businesses,” said Sam Altman, co-founder and CEO of OpenAI. “Developing our own accelerators adds to the broader ecosystem of partners all building the capacity required to push the frontier of AI to provide benefits to all humanity.”
Broadcom’s CEO, Hock Tan, framed the collaboration as a historic milestone. “Broadcom’s collaboration with OpenAI signifies a pivotal moment in the pursuit of artificial general intelligence,” Tan stated. “OpenAI has been in the forefront of the AI revolution since the ChatGPT moment, and we are thrilled to co-develop and deploy 10 gigawatts of next- generation accelerators and network systems to pave the way for the future of AI.”
This venture signals a major industry shift where leading AI software creators are now vertically integrating into custom silicon to overcome the limitations of off-the-shelf hardware. For Broadcom, the partnership solidifies its position at the forefront of providing the essential building blocks for the world’s most demanding AI workloads, combining custom accelerators with open, scalable networking to define the next generation of AI infrastructure.
