OpenAI and Broadcom unveil LLM-optimized inference chip
中文摘要
OpenAI与Broadcom推出Jalapeño芯片，专为LLM推理优化，以提升AI性能与效率。
English Summary
OpenAI and Broadcom unveil Jalapeño, a custom chip optimized for LLM inference to improve AI performance, efficiency, and scale.
原文节选
OpenAI and Broadcom introduce Jalapeño, a custom AI chip built for LLM inference to improve performance, efficiency, and scale across AI systems.