OpenAI Used Its Own LLMs to Design the Jalapeño Chip in Under 20 Months
How OpenAI Used Its Own LLMs to Design Its Jalapeño Chip
OpenAI's first AI accelerator, Jalapeño, went from concept to silicon in under 20 months, with only nine months from first RTL to tape-out. A team of fewer than 100 engineers used internal LLMs and the XLS high-level synthesis chain to accelerate front-end design, while Broadcom handled physical implementation. The chip delivers 13.4 petaflops of 4-bit compute and 232 GB of HBM4 memory at 15.4 TB/s. Experts call the timeline best in class, but note Broadcom's role was essential.
The models are giving superpowers to our engineers. Our engineers are still driving the work. They're still the final arbiter of what's going on. But they can do things a lot faster. They can explore a lot more paths.