Why is OpenAI developing its own chip?

To reduce reliance on external chip suppliers like NVIDIA and improve the performance and efficiency of its models.

Who are the partners in developing the chip?

Broadcom (AVGO) for silicon technology and Celestica for rack technology.

When will the chip be available?

OpenAI has not announced availability yet.

OpenAI Unveils Custom 'Jalapeño' AI Inference Chip

OpenAI has launched the 'Jalapeño' custom AI inference chip, built in collaboration with Broadcom and Celestica, as part of its push to enhance AI infrastructure.

June 24, 2026

2 min read

Source: Barrons.com

OpenAI, the creator of ChatGPT, has announced the launch of its custom AI chip named 'Jalapeño', specifically designed for inference tasks. The chip was developed using silicon technology from Broadcom (AVGO) and rack technology from Celestica.

The Product

The 'Jalapeño' chip is designed to accelerate inference in generative AI models, reducing latency and improving energy efficiency. This move comes as OpenAI seeks to reduce reliance on external chip suppliers like NVIDIA (NVDA).

Pricing and Availability

OpenAI has not yet disclosed pricing or availability details. The chip is expected to be used internally to power ChatGPT and other models.

Competition

This announcement is part of a broader tech race among industry giants to develop custom AI chips. Microsoft (MSFT), Amazon (AMZN), and Google (GOOGL) are all pursuing their own chip designs to cut costs and boost performance.

Potential Impact on the Company

The 'Jalapeño' chip could enhance OpenAI's competitive edge and lower operational costs in the long run, but it requires significant R&D investment.

Frequently Asked Questions

It is a custom AI inference chip developed by OpenAI in collaboration with Broadcom and Celestica.

Found this useful? Share it

This article was rewritten in Wrqti's editorial style based on information from the original source above. Content is informational only — not investment advice.