OpenAI's first chip, 'Jalapeño', went from design to silicon in nine months (techcrunch.com)
The part targets inference, not training, so Nvidia keeps the training moat while OpenAI chases the recurring per-token cost. The pitch leans on one phrase: development "accelerated by OpenAI's models".
On Hacker News a chip engineer called the nine-month timeline impossible and guessed it is a relabeled Broadcom design with little real OpenAI contribution.
Sign in to save