OpenAI and Broadcom unveil Jalapeño chip for scalable LLM inference
OpenAI and Broadcom have developed Jalapeño, an ASIC optimized for large language model inference, promising improved performance per watt and marking the start of a long-term AI chip roadmap.
