OpenAI, Broadcom unveil chip to run models faster, cheaper

Published Wed, Jun 24, 2026 · 11:07 PM

OpenAI unveiled its first custom artificial intelligence chip developed in partnership with Broadcom, part of a bid by the ChatGPT maker to gain an edge by tailoring hardware to better run its AI products. 

The companies said on Wednesday (Jun 24) that OpenAI has received the first samples of the chip, called Jalapeno, and is testing how the silicon handles running AI workloads.

So far, the accelerator is showing cost savings of roughly 50 per cent compared with typical AI graphics processing units, Broadcom chief executive officer Hock Tan said in an interview. 

The finalised chips will be integrated into large data centres from OpenAI-backer Microsoft and other partners starting later this year.

Tan expects OpenAI and Broadcom should be able to exceed his prior forecast for deploying 1.3 gigawatts’ worth of chips next year. “We like to think we can do better because there is a lot of demand,” he said. 

Though OpenAI continues to rely heavily on chips from market leader Nvidia, the startup has worked to broaden its mix of suppliers to meet surging demand for its AI services. OpenAI has struck multibillion-dollar deals with chipmakers such as Advanced Micro Devices and Cerebras Systems.

Shares of Broadcom gained 1.6 per cent to US$386.25 at 9.35 am in New York trading. They were up almost 10 per cent this year through Tuesday’s close.

In October, OpenAI announced that it would partner with Broadcom to design accelerators tailored to work best with its artificial intelligence models.

OpenAI plans to spend tens of billions of dollars on Broadcom chips, Bloomberg News has reported, adding to the unprofitable startup’s massive outlays on physical infrastructure to support AI. 

OpenAI’s head of hardware, Richard Ho, declined to say how the company will pay for the chips, noting that financing arrangements would be finalised when the full order of processors is complete.

Tan declined to comment, but reiterated that Broadcom has set up a chip financing vehicle with Apollo Global Management and Blackstone that will also aid OpenAI. 

OpenAI raised US$122 billion in funding earlier this year to bolster its costly push for chips, data centres and talent. The company has also entered into financing deals with suppliers, including Nvidia and AMD, that have been criticised for being circular in nature. 

The new chips were developed from scratch in record time, according to OpenAI and Broadcom. OpenAI said it is still assessing the test versions but early data shows Jalapeno is capable of delivering performance per watt of energy that’s “substantially better than current state-of-the-art.” 

The chip was designed for large language models, or LLMs, and guided by the artificial intelligence startup’s approach for inference — a stage that involves running AI tasks — Ho said. Still, he said, “it is a very general purpose device” that “will address future LLM innovations.” 

OpenAI wants to keep its options open as to whether it would let other model makers use its chips, Ho said. 

Within the tech industry, some AI chips are more powerful for doing many tasks at once while others are focused on providing a single answer to a user as fast as possible.

OpenAI said its goal is to combine both — the power of leading AI chips with speed that’s closer to specialised hardware for running models faster. 

The chips are meant to perform better by reducing the amount of data that must move around. The hardware is also designed around the usage patterns of computing, memory and networking gear that are most important for top-of-the-line AI models. 

The two companies have a road map for future chips, with the next version planned for 2028 and then annually after that, Tan said. While Jalapeno is focused on inference, OpenAI may consider other workloads with future chip generations. 

For OpenAI, the chip represents the start of a series of steps to exert control over all levels of the infrastructure it relies on, Ho said. Tan, meanwhile, expects other AI developers to follow in OpenAI’s footsteps. 

“At the end of the day, every one of the frontier model creators — and there are not that many of them at this point outside of China — but every one of them, over time, will create their optimum, custom AI accelerator and networking,” Tan said. BLOOMBERG