OpenAI and Broadcom unveil new chip tailored for large-scale LLM inference

OpenAI and Broadcom have announced a new chip called Jalapeño, designed specifically for large language model inference in data centers. The chip is intended for deployment at scale, with both companies describing it as the first generation in a long-term project that will see continued refinement. This partnership aims to optimize hardware for the growing demands of AI inference workloads.