6 ms·OpenAI and Broadcom announce chip designed for LLM inference at scale2 points by rajsuper123 3mo ago