OpenAI and Broadcom Launch Jalapeño, a Revolutionary AI Processor

OpenAI and Broadcom Launch Jalapeño, a Revolutionary AI Processor

OpenAI recently unveiled a new intelligence processor for AI servers, developed in partnership with Broadcom. This processor, named Jalapeño, is designed to power current and future large language models, according to an announcement made on Wednesday.

Jalapeño is an ASIC (Application-Specific Integrated Circuit), meaning it is designed for a specific purpose: AI inference. With AI inference, models process a user's request to execute an agent like Codex or provide a response from ChatGPT, while AI training involves a model consuming vast amounts of data to inform its responses.

This launch comes just nine months after OpenAI announced its collaboration with Broadcom to create its own chips, in order to reduce the company's reliance on GPUs provided by Nvidia, which are in limited supply. In an interview with Reuters, Broadcom CEO Hock Tan stated that Jalapeño will match the performance of Nvidia's Blackwell chips and Google's Tensor processing units. Microsoft, Meta, and Amazon are among other AI companies that have also recently launched custom-designed AI chips to power their servers, whether for training or inference, while still lagging behind Nvidia's chips in terms of overall performance.

OpenAI views Jalapeño as the "first step in a multi-generational computing platform," which it plans to roll out by the end of 2026. "Although OpenAI is still measuring the final performance, early tests show that Jalapeño will offer a substantially better performance-per-watt ratio than the current state of the art," says OpenAI.