OpenAI and Broadcom unveil LLM-optimized inference chip

OpenAI and Broadcom launched Jalapeño, a custom chip for running large language models (LLMs) more efficiently.

מקור: OpenAI News — לכתבה המלאה