Keep up the great work! This is a fantastic initiative of the zcash team. Quite important for a fair Zcash launch and for Zcash transparency, openness and decentralization. Keep it up!
Also thank you to Zooko and Least Authority for the open bounty.
So, how much for a CPU miner (equihash solver) that is:
-Standalone plain C implementation with no extra dependencies
-20%-30% performance improvement
-Stable memory usage, no spikes to 2x the necessary since the memory allocation is explicitly controlled (can mine 2 cores on 2Gb of RAM, whre zcashd crashes)
A couple of us at Toomim Bros Bitcoin Mining Concern are interested in making a GPU miner for Equihash/ZCash. We started researching the algorithm a few days ago. Weāre targeting OpenCL on AMD.
For the sake of public record, here is the SHA256 hash of the tarball with the code of my CPU miner - 1e63e4fab651e860803782e282e4d2c4e2539b56318c9381d631a908e0ccae07
This is posted in the spirit of ābe preparedā, in case iāll enter the contest and it would involve time-sensitive conditions or any other need to prove the authorship of that code.
@prometheus We have a few hundred AMD GPUs, so weāll be focusing on those. OpenCL does run on NVIDIA hardware, though not very well, so in theory people with NVIDIA GPUs should be able to get at least some hashrate out of our miner. Maybe after version 1.0 we can look into adding CUDA support.
@nimos Definitely Linux, possibly Windows as well.
Weāre not very far along yet. Right now Iām working on writing a python implementation of Equihash just to make sure I understand the algorithm. I suppose that technically might qualify as a standalone CPU miner, but we wonāt be submitting it for this contest because I think it would exceed the permissible limit on silliness for the project.
According to the timeline, it looks like a winner wonāt be announced until Dec 2nd. Does that mean we wonāt have the open source CPU/GPU miner until then? I was hoping for one at launch.
Interesting to see this positioned as a cross-pollination contest:
entries are rewarded both for having their ideas incorporated into others, and for incorporating otherās ideas.
Do we know anything about the relative value of these two rewards?