I’ve got 2 970 PROs in RAID0 (total 2TB) and 4 Intel s4610 in RAID0 as well (also total 2TB), 5950x and 64GB of 3600Mhz RAM.
I’m using Swar plot manager and have the following config:
2 jobs - one for the Samsung temp and one for the Intel temp. Each has - 9 in parallel, 10 with start early, 4 at phase 1, 3800 RAM, 4 threads, 45mins of stagger, 128 set as a priority in windows.
For the past 10-12 days I’ve been noticing that the plots from the Samsung job are getting slower and slower and now are at a point where a Samsung plot takes more than 11:30 hours, while an Intel plot takes about 10:50 hours. They’ve both went up from something like 9 - 9:30 hours form two weeks ago. I notice that Phase1 and Phase3 in particular has went up by more than half an hour each (almost 4:30 for Samsung…). Do I overload my CPU? I have a good cooler and it doesn’t seem to throttle although it stays at 100% most of the time. Also, on one of my Samsung Nvmes the Temp2 in HwInfo stays at about 90 degrees which I believe is a lot. Could the drive throttle or something?
The previous setup was the same but with only 6 plots in phase 1 in total so 3 per job but the timings weren’t much better and were slowly climbing day-by-day…
Here’s a screenshot of the swar at the moment. You can see how the Intel jobs take over the Samsung ones…
Also, I’ve just noticed that “start early” plots have fired for both jobs, although only one is at phase 4. Does someone know why this may be?
Any insights?