AI in the Field

AI in the Field · Oct 9, 2026

AI wrote everything from chip RTL to compiler, hitting 85 tokens per second on a used FPGA of about $300. What set the speed was its pass/fail verification

Independent developer FeSens had AI write everything from the hardware design to the compiler, then ran a 230M model at 85.8 tokens per second on a used FPGA. The speed came from a verification setup that passes or rejects each proposed change. The project does not include comparison figures, and it does not say how much a human was involved.