I run a small experiment: five Bitcoin Core forks merged-mined with RandomX on ordinary CPUs (one hash attempt mines all five chains). Mainnet started on 2026-11-06; announcement thread with all details:
https://bitcointalk.org/index.php?topic=5599778.0I'd like the view of people who have mined small CPU coins, because the first two weeks show a pattern I did not expect this early.
What the data shows (live:
https://sentodaimon.github.io/coin-series/)
- Week 1: only my desktop, about 1,300 H/s.
- About 20 minutes after the announcement outside miners arrived; within a day the network peaked near 60,000 H/s. Per-block ASERT difficulty caught up and block intervals went back to about 10 minutes. No reorganizations so far.
- Since then the hash rate has drifted down to 13,000-20,000 H/s, and the number of outside reward addresses in the last 100 blocks went from 5 to 2.
- Share of the largest address in the last 100 blocks: 48% (Nov 19), 87% (Nov 20), 53% (Nov 21), 69% (Nov 22). My own desktop finds 5-11%.
What I have not done, on purpose- No emergency rule changes. Changing consensus because one miner is winning would hurt trust more than the concentration does.
- No paid promotion, no bounties, no listing. There is nothing to sell.
Questions- On small RandomX coins you have mined, did concentration like this fade as more people joined, or does it usually stay?
- Is the bigger risk a single miner doing something hostile, or simply leaving and taking most of the hash rate at once? (Stall relief eases difficulty after an hour without a block, so the chain should recover, but I would like to hear real experiences.)
- Anything that would make it easier for you to point a few CPU threads at it? Current setup: Windows zip (node + miner), or build from source on Linux with the patches in the repo.
Disclosure: I am the developer. No premine, no sale, no promised value.