GitHub openTPU Ships AI-Designed FPGA Inference Accelerator Running Qwen3 and LFM2.5 on a Kintex-7

GitHub user FeSens published openTPU, a self-described 'AI-developed' inference accelerator spanning the full stack — RTL, ISA, simulator, compiler and host software — targeting a Xilinx Kintex-7 FPGA PCIe card and supporting Qwen3, LFM2.5 and Qwen3.5 inference. The project explicitly tests whether AI agents can design the chip that runs their own inference, extending the earlier auto-arch-tournament experiment. The repo trended on Hacker News with 154 points and 161 comments.