EventDevelopersAugust 25, 2026

OpenAI's Jalapeño chip beats Nvidia Blackwell on inference benchmarks

At Hot Chips, OpenAI shared benchmark results for its Jalapeño inference chip, showing more tokens per user and more throughput per kilowatt than Nvidia Blackwell on SemiAnalysis's InferenceX benchmark. Deployment begins in small volumes at end of 2026, scaling in 2027.

Featured · Richard Ho

1 source

More stories today

Open the live feed