So rn I have a 24 core server. 180gb ram. And A Gtx 960. And a 1 tb ssd and many tb.
Does this work???
How exactly does one get started with gpu farming/ plot compression.
What would the optimal config be?
What clients?
I have my tbs filed rn. But I’m willing to just replot.
Start a fresh.
This is a good start - chia-gigahorse/cuda-plotter at master · madMAx43v3r/chia-gigahorse · GitHub
k32 / C7/8 need 198 GB or so RAM for full RAM plotting, so you are just a tad under. 980 will be potentially useless, but there is no reason not to test it. Not sure how much RAM it has, as that could be a bummer. The assumption is that if you are using GPU, you only need a handful of cores, so you may want to depopulate your box to save on power.
The rest really depends what you have on hand, and you didn’t provide any info about what you have.