Notes from building Strike a Block: repairing and driving one USB Bitcoin ASIC, measuring it, and the Bitcoin block format it imitates. Written by Laksh Gandikota from the project's own code, logs and commit history.
-
I fixed a dead Bitcoin ASIC and wrote the driver from scratch
A Compac F that answered but would not hash, a trim pot at 1.398 V, and a Python driver whose worst bug was 9-byte frames split across USB reads.
-
From 78 to 137 GH/s by changing one number
The work interval said 22 ms and the log said 32. Benchmarks, a 30-minute soak, and why no trick gets you more leading zeros.
-
The genesis block message, byte by byte
The headline is not in the 80-byte header. Here is the field it is in, the exact offset, and how the header still locks it down.
-
How long it takes to mine leading zeros
16 to the n, a table at three hash rates, why the average misleads, and where today's Bitcoin blocks sit.
-
Messages mined into Bitcoin blocks
Seven documented coinbase messages from block 0 to block 840,000, each read from the raw bytes and linked to its block.
Also: Genesis Block Day, a page about 3 January and block 0. The project itself is on the home page.