OpenAI has announced its first custom-built artificial intelligence chip, marking a significant step in its efforts to expand beyond consumer-facing products and strengthen its position in AI infrastructure.
The new processor, named Jalapeño, was developed in partnership with Broadcom and is designed to handle the computing demands of ChatGPT and OpenAI's coding assistant, Codex, Caliber.Az reports via US media.
The company says the chip is also intended to support future generations of AI systems.
In a blog post announcing the launch, OpenAI said: “While OpenAI is still measuring final performance, early testing shows that Jalapeño will deliver performance per watt substantially better than current state-of-the-art.”
The release follows an announcement last year that OpenAI and Broadcom would work together to develop custom chips capable of powering 10 gigawatts of computing capacity. Wednesday's announcement marks the first product to emerge from that partnership.
Technology firms are increasingly investing in their own AI processors as they seek to reduce dependence on external suppliers, particularly chipmaker Nvidia, whose hardware dominates much of the AI industry.
Custom chips can be tailored to specific AI workloads, potentially improving efficiency and lowering costs. OpenAI said Jalapeño was built specifically for modern large language models rather than as a general-purpose processor.
The move comes as OpenAI prepares for a potential initial public offering that could reportedly value the company at around $1 trillion, increasing pressure on the firm to demonstrate sustainable revenue growth.
At the same time, competition among AI companies has intensified as firms race to secure the computing power and energy resources needed to support increasingly sophisticated AI services. OpenAI has previously highlighted the need for substantial investments in infrastructure to reduce the cost of operating advanced AI systems.
The company says the new chip could help make its services more affordable and accessible by improving efficiency across its computing operations.
Commenting on the announcement, OpenAI co-founder and president Greg Brockman said: “By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access.”
The launch places OpenAI alongside major technology companies such as Google and Amazon, both of which have invested heavily in developing proprietary AI chips as demand for computing power continues to surge.
By Aghakazim Guliyev
