I have been working for the past couple of weeks on a miner which significantly outperforms existing options. It contains both systems improvements and better algorithms/data structures derived from my research. My miner is currently approx. 50% faster than the state-of-the-art, and I think I can get another 50% on top of that with a bit more work. At this point, my miner is only for CPUs. It could be ported to run on GPUs given time (not sure what the perf improvements would be).
The purpose of my post is twofold.
FIrst, I want to gauge interest in CPU miners. Is anybody currently mining using CPUs? If so, which CPUs, architecture, cores, frequency, etc. and how much RAM and bandwidth?
Second, I want to get an idea for what would be an acceptable “release model” in this community? I was thinking of releasing a binary with a small dev fee, that could possibly be turned off with a one-time donation to support ongoing development. Are people opposed to that model? If so, how else should it be done? I am not looking to get rich or anything, but I think it is only fair to get a small reward in exchange for long hours of development In any case the code will eventually be open sourced once I publish the paper I am currently writing.
i7 5820K @ 49S/s using 11 threads with quad channel ddr4 3200 16 Gb if we had a miner that could get 100 S/s on CPU I would start her up again win 10 x64
i7 6800k at 4.0 ghz with 11 threads and dual channel 2400mhz 16gb ram = 30 sols/s
A dev fee sounds fair to me and I will definitely try out your miner.
With gpu mining pulling so far ahead, you’ll need to offer compelling speed and efficiency otherwise most people will be losing money to electricity.
i stopped CPU mining around a week ago. it was not profitable anymore. what’s most important i think is the amount of solls per watt ratio. if that’s a profitable number… you will have fans for sure
In any case the code will eventually be open sourced once I publish the paper I am currently writing.