XCHSystems

> > > The -r flag is just a multiplier, even if you set it low it's going to try to use available threads. I've found that phase 2 does...

I just created two plots in parallel, both plots took 15 minutes to create, so that's two plots in 15 minutes, therefore in theory I should be able to get...

> > > Could the PMEM be your IO bottleneck? Can you detach the 256gb RAM from it and create a 220G tmpfs with it? Edit: Actually I guess since...

I do not think that's the problem here, as I know I can push I/O to the storage without SPDK using FIO, and from the iostat screen the I/O being...

> > > I think your setup/statement is wrong. If you are talking about luxury hardware and records, the record is ~60 seconds on a computer cluster, this uses 384...

 This is the 512GB RAM Disk test, I am currently performing the same tests with a single P5800X, and then again with a P5800X+128GB RAMDisk

To verify what @gryan315 was saying, I ran two instances of MadMax using 28 threads each, both plots completed in 15 minutes, which means that the bottleneck is not hardware...

I have created thousands of plots with bladebit_cuda and they all pass the plot check

@Jsewill Tell us more about your system, what GPU, CPU, RAM etc, where you are storing your plots temporarily, where your final destination is. For example in my system Xeon...

> > @Jsewill Tell us more about your system, what GPU, CPU, RAM etc, where you are storing your plots temporarily, where your final destination is. For example in my...