OpenAI 官方动态：OpenAI and Broadcom unveil LLM-optimized inference chip
原文摘要：OpenAI and Broadcom introduce Jalapeño, a custom AI chip built for LLM inference to improve performance, efficiency, and scale across AI systems. 来源：OpenAI 官方动态。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。