Jalapeño’s first results show industry-leading speed and efficiency in AI inference
OpenAI's custom Jalapeño chip delivers industry-leading AI inference speed and efficiency.
OpenAI has unveiled Jalapeño, its first custom inference chip, claiming industry-leading throughput and lower latency for modern AI models. This marks OpenAI's entry into custom silicon, reducing dependence on third-party hardware like Nvidia. It signals a broader trend of frontier AI labs vertically integrating into chip design to control inference cost and speed.