OpenAI to Launch Its First AI Chip with Broadcom by 2026? Check here!

Sep 5, 2025, 09:00 EDT

OpenAI is reportedly partnering with Broadcom to develop its first AI chip, aiming for a 2026 release. This strategic move seeks to reduce reliance on Nvidia's expensive and scarce GPUs, cut costs, and enhance control over AI performance. The custom chip will be used internally by OpenAI for training its models, following a trend set by other tech giants like Google, Amazon, and Meta. TSMC is expected to handle manufacturing.

Open AI
Open AI

The AI race is heating up, and OpenAI, the company behind ChatGPT, is stepping into a whole new game. Reports suggest that OpenAI is working on its very first artificial intelligence chip, in collaboration with Broadcom, and could roll it out as early as 2026. Let’s understand this all in simple language.

Check out: Is OpenAI ChatGPT Down? Here are the Best AI alternative chatbots - Google Gemini, Perplexity, Meta AI, and Others

Why is OpenAI Making Its Own AI Chip?

If you’ve used ChatGPT or seen AI tools in action, you already know they need massive computing power. Right now, most AI companies, including OpenAI, rely heavily on Nvidia’s GPUs (graphics processing units) to train and run their systems. But here’s the catch: demand for these chips is exploding worldwide, prices are high, and supply is tight.

By building its own chip, OpenAI can:

  • Cut costs in the long run

  • Depend less on Nvidia and other chip suppliers

  • Have more control over performance and innovation

What Role is Broadcom Playing in This?

Broadcom is a semiconductor giant in the U.S. and is partnering with OpenAI, it appears, to engineer and develop this chip. Broadcom CEO, Hock Tan, hinted that one of their big new customers ordered $10 billion of AI infrastructure, though he never named OpenAI. 

This means OpenAI is not dabbling or experimenting - it's legitimately trying to seize control of its hardware future.

Will This New Chip Be Available to Everyone?

Not really. Reports suggest that OpenAI plans to use this chip internally for training and running its own AI models rather than selling it to other companies or the public. In short, it’s not like buying a graphics card for your PC; this is all about powering the brains behind ChatGPT and future OpenAI products.

How Does This Compare to Other Tech Giants?

OpenAI isn’t the first to go down this road. Rivals like Google, Amazon, and Meta have already created custom chips for their immortal AI workloads. Their custom-designed chips are designed to process the AI workload faster, cheaper, and more competitively for specific tasks. 

So OpenAI appears to first catch up and possibly to prepare to compete with these companies.

Check out: AI Psychosis Explained: Symptoms, Causes, and What You Need to Know

What’s Next

The reports suggest that OpenAI is still in the process of finalizing the design of its first chip. Once ready, fabrication (the actual manufacturing) will likely be handled by TSMC (Taiwan Semiconductor Manufacturing Company), one of the biggest chipmakers in the world.

If everything goes smoothly, we could see OpenAI running on its very own custom-built chips by 2026.

Sneha Singh
Sneha Singh

Content Writer

    Sneha Singh is a US News Content Writer at Jagran Josh, covering major developments in international policies and global affairs. She holds a degree in Journalism and Mass Communication from Amity University, Lucknow Campus. With over six months of experience as a Sub Editor at News24 Digital, Sneha brings sharp news judgment, SEO expertise and a passion for impactful storytelling.

    ... Read More

    Trending Tags