SAN FRANCISCO (Diya TV) — OpenAI and Broadcom announced a major collaboration to deploy 10 gigawatts of custom AI chips designed to meet the world’s growing demand for artificial intelligence. The partnership combines OpenAI’s expertise in AI model development with Broadcom’s advanced networking and semiconductor technology.
By designing its own AI accelerators and partnering with Broadcom for development and deployment, OpenAI aims to push the boundaries of AI performance. The new hardware will be used across OpenAI’s facilities and in partner data centers worldwide.
OpenAI’s custom chips will embed insights from its experience creating advanced AI models. By integrating these lessons directly into hardware, OpenAI expects to unlock higher levels of AI capability and intelligence. This move also strengthens OpenAI’s broader ecosystem of partners and helps scale AI infrastructure globally.
“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 building the capacity required to push the frontier of AI and provide benefits to all humanity.”
Broadcom will provide the networking solutions that allow OpenAI’s AI clusters to scale efficiently. The collaboration includes deploying racks using Broadcom’s Ethernet, PCIe, and optical connectivity systems. These technologies support scale-up and scale-out networking, essential for high-performance AI applications.
“Broadcom’s collaboration with OpenAI signifies a pivotal moment in the pursuit of artificial general intelligence,” said Hock Tan, president and CEO of Broadcom. “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.”
Charlie Kawwas, president of Broadcom’s Semiconductor Solutions Group, emphasized the combination of custom AI chips and Ethernet networking. “Custom accelerators work remarkably well with standards-based networking solutions,” he said. “This approach provides cost-optimized, high-performance AI infrastructure that sets new industry benchmarks.”
OpenAI has grown to more than 800 million weekly active users. Its AI services are widely adopted by global enterprises, small businesses, and developers. The partnership with Broadcom aims to meet surging demand for AI across industries and applications.
By producing its own chips and systems, OpenAI can directly apply insights from developing advanced models to hardware design. This integration is expected to improve efficiency, speed, and intelligence in AI systems.
Greg Brockman, co-founder and president of OpenAI, said the collaboration will accelerate breakthroughs in AI. “By building our own chip, we can embed what we’ve learned from creating frontier models and products directly into the hardware, unlocking new levels of capability and intelligence,” Brockman said.
The new racks will feature Broadcom’s end-to-end Ethernet and connectivity solutions. These systems support high-performance, scalable AI clusters that are also energy-efficient. For Broadcom, the partnership highlights the value of custom accelerators and Ethernet-based networking for AI data centers.
