OpenAI has announced a new partnership with technology giant Broadcom to build powerful computer chips that will make artificial intelligence (AI) work faster and smarter.
Both companies have signed an agreement to produce racks (large sets of computers) that will use Broadcom’s high-speed networking technology to connect the AI systems.
OpenAI will design the accelerators and systems, which will be developed and deployed in partnership with Broadcom. The goal is to create enough power, 10 gigawatts worth, to support OpenAI’s growing AI tools and products.
The project will begin in 2026 and is expected to be completed by 2029.
What they said
Sam Altman, co-founder and CEO of OpenAI, described the partnership as a major milestone in advancing AI infrastructure.
“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.
“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,” Altman said.
“Broadcom’s collaboration with OpenAI signifies a pivotal moment in the pursuit of artificial general intelligence. 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,” Tan said.
According to Tan, the collaboration showcases how semiconductor innovation and AI research can converge to push the limits of computing performance.
Greg Brockman, co-founder and President of OpenAI, explained that the collaboration with Broadcom will help achieve major progress in AI performance and accessibility. He noted that building OpenAI’s own chips allows the company to apply lessons learned from developing advanced models directly into hardware, creating faster, smarter, and more capable AI systems.
The move is expected to make OpenAI one of the few companies that design both large-scale AI models and the specialized chips needed to power them efficiently.
Charlie Kawwas, President of Broadcom’s Semiconductor Solutions Group, said the partnership sets new standards for building scalable and energy-efficient AI systems.
He explained that custom accelerators work seamlessly with Broadcom’s Ethernet networking technology to create cost-effective and high-performing AI infrastructure.
The racks will include Broadcom’s full range of Ethernet, PCIe, and optical connectivity tools, reinforcing its leadership in AI infrastructure development. Broadcom’s main focus will be on deploying efficient Ethernet-based networking systems designed to handle massive AI workloads with high performance and reliability.
OpenAI has been on an aggressive expansion streak this year, securing major funding and strategic acquisitions to strengthen its position in the AI industry.
OpenAI has been on an aggressive expansion streak this year, securing major funding and strategic acquisitions to strengthen its position in the AI industry.
The company also signed a $300 billion cloud computing deal with Oracle to secure the power needed to train and run large AI models. The agreement, spanning five years, involves around 4.5 gigawatts of computing capacity.