OpenAI and Broadcom announce chip designed for LLM inference at scale
Summary
OpenAI and Broadcom announced a new computer chip named Jalapeño, made to run large language models (LLMs) like ChatGPT more efficiently in data centers. The chip is designed specifically for this task and is expected to be used in data centers by the end of the year, with ongoing testing to confirm its performance.Key Facts
- The chip is called Jalapeño and is made for large language model inference, which means it helps run AI models more efficiently.
- It is an ASIC, a special chip built for one specific use rather than general tasks.
- Broadcom designed the chip based on detailed information from OpenAI about their future AI models.
- The chip development took nine months from design to production.
- OpenAI says early tests show Jalapeño uses less power for better performance compared to current chips.
- This chip aims to reduce reliance on other chipmakers like Nvidia by integrating more of the technology in-house.
- Jalapeño will help data centers handle more computing power during a time when demand for AI computing is very high.
- Both companies plan to have the Jalapeño chips running in data centers before the end of this year.
Read the Full Article
This is a fact-based summary from The Actual News. Click below to read the complete story directly from the original source.