OpenAI has unveiled ‘Jalapeño,’ a custom chip designed specifically for large language model inference. The chip, developed with Broadcom, promises better performance per watt and faster running times. This move marks OpenAI’s entry into custom hardware, which could make running AI models cheaper and more reliable. Why it matters: By controlling the full stack from chip to product, OpenAI can potentially reduce costs and increase the reliability of its AI models, a significant development in the pursuit of more widespread AI adoption.