OpenAI says its Jalapeño chip delivered 1.5 to 1.9 times more AI work per watt than Nvidia's GB200 and GB300 superchips on the InferenceX benchmark, tested across GPT-OSS 120B, DeepSeek R1, and Kimi K2.5 1T. The Broadcom-made ASIC also showed 1.7 to 3.6 times lower latency. Hardware VP Richard Ho said OpenAI will deploy the chip in small volumes this year, ramping up through 2027, while continuing to work with Nvidia.
No score is assigned. Sources and their independence are shown in the citation chain below.