OpenAI and Broadcom Partner to Build 10 Gigawatts of Custom AI Accelerators: A New Era of AI Hardware Begins

In one of the most ambitious tech collaborations of the decade, OpenAI and Broadcom have announced a multi-year partnership to co-develop and deploy 10 gigawatts of OpenAI-designed AI accelerators, marking a monumental shift in how the world’s most advanced artificial intelligence systems will be powered and scaled.

The companies revealed that OpenAI will design the accelerators and systems, while Broadcom will develop and deploy racks of these custom-built AI systems, fully equipped with Ethernet and networking solutions from Broadcom. The rollout will begin in the second half of 2026, with full deployment expected by 2029.

Strategic Alliance to Power the Future of AI

This collaboration marks OpenAI’s official entry into chip design, signaling its move toward greater vertical integration. By designing its own hardware, OpenAI aims to directly embed learnings from its frontier AI models and products into the chips themselves — an innovation that could redefine performance and efficiency in AI computation.

“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, OpenAI’s co-founder and CEO. “Developing our own accelerators adds to the broader ecosystem of partners building the capacity required to push the frontier of AI.”

Sam Altman of OpenAI

For Broadcom, the deal underscores its role as a key infrastructure enabler for the global AI boom. “OpenAI has been at the forefront of the AI revolution since the ChatGPT moment,” 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.”

OpenAI and Broadcom will co-develop systems that include accelerators and Ethernet solutions from Broadcom for scale-up and scale-out.

Learn more: https://t.co/btVN2xLIkg pic.twitter.com/MpBdhZUvsn

— Broadcom (@Broadcom) October 13, 2025

Road to Artificial General Intelligence

OpenAI President Greg Brockman emphasised that this move will help the company advance toward its ultimate goal: artificial general intelligence (AGI). “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,” he said.

The AI accelerators will be housed in Broadcom-developed racks that utilize standards-based Ethernet, PCIe, and optical connectivity — offering a cost-optimized, power-efficient, and scalable infrastructure. This design approach could set a new industry benchmark for AI datacenters, combining custom compute power with open networking standards.

Mattel Partners with OpenAI to Test Sora 2, the Next-Gen AI Video Model Revolutionizing Toy Design

Scaling AI for Billions

OpenAI, which now boasts over 800 million weekly active users, says the partnership will help meet surging global demand for AI applications across enterprises, small businesses, and developers.

“This collaboration will power breakthroughs in AI and bring the technology’s full potential closer to reality,” Brockman said.

Leap Forward for AI Infrastructure

Industry analysts are calling the collaboration a defining moment for the AI hardware ecosystem, as companies increasingly seek in-house chip solutions to reduce dependence on third-party GPU suppliers. The decision to deploy Ethernet-based clusters also highlights a shift away from proprietary networking systems, potentially making large-scale AI clusters more open, flexible, and energy-efficient.

As AI workloads grow exponentially, OpenAI’s partnership with Broadcom is more than a hardware upgrade — it’s a declaration of intent. The race to AGI now runs not just on algorithms, but on the silicon that powers them.