OpenAI Develops Custom AI Chip 'Jalapeño' with Broadcom
OpenAI announced plans to develop a custom inference chip codenamed Jalapeño in collaboration with Broadcom, joining a growing list of companies building their own chips to reduce dependence on Nvidia.
OpenAI has announced plans to develop a custom AI inference chip codenamed Jalapeño, in collaboration with Broadcom (AVGO). The move aims to reduce single-supplier risk, a clear reference to Nvidia's dominance in the AI chip market.
The Product
Jalapeño is an inference chip designed to run pre-trained AI models more efficiently and at lower cost. OpenAI is partnering with Broadcom for manufacturing, leveraging Broadcom's expertise in custom chip design.
Pricing and Availability
OpenAI has not disclosed pricing details or a commercial launch timeline. The chip is expected to be used internally first before being offered to customers.
Competition
This move comes as major companies like Google (GOOGL), Apple, and SpaceX race to develop their own chips, increasing pressure on Nvidia, which dominates the AI chip market. OpenAI currently uses Nvidia chips for training its models.
Potential Impact on the Company
Developing a custom chip could lower operating costs in the long run and improve the performance of OpenAI's models, while reducing the risks of single-supplier dependency. It also strengthens Broadcom's position in the custom chip market.
Frequently Asked Questions
Found this useful? Share it