My little $40 FPGA can beat your big bad GPU. Prove me wrong and win 0.1 BTC. I designed a proof-of-work called BRAM-ring. A $40 Efinix Tz170 does 37,100 H/s on it. The best anyone has managed on a GPU is 1,493 H/s, on a $499 RTX 3070. Per dollar the little chip wins by more than 100×, and I don't think any GPU, CPU, NPU or other off-the-shelf hardware can close that gap. The bounty: • 0.1 BTC to the first person who shows non-FPGA hardware at or above 371 H/s per dollar • 0.01 BTC every time someone beats my GPU or CPU kernel by 10% (bit-exact, measured on my machine, the bar ratchets) • 0.01 BTC for a real shortcut in the function: half the work per hash, counted not timed How to claim: reply to this tweet with the full 40-character git commit hash of your code and the prize you're claiming. Your code can stay private until the deadline; the hash fixes your place in line. Deadline: 4 November 2026, 23:59 Pacific. Rules: git.nodelabs.ca/nexapow/bram… Code (make prove checks C, AVX2, CUDA and Verilog against the same 8 vectors): git.nodelabs.ca/nexapow/bram… Prize pool 0.2 BTC, held at bc1qtk8q5nsupm47kg5vz9turqacx8lg2whvxukpuj. Signed proof of funds in the first reply and in BOUNTY.md.