OpenAI and Broadcom unveil LLM-optimized inference chip
=======================================================

OpenAI and Broadcom introduce Jalapeño, a custom AI chip built for LLM inference to improve performance, efficiency, and scale across AI systems.