Hi all, just thought I would get a quick sanity check on my FCLK's stability.
I was previously running 1.15VSOC, 1.15VDDP, and 910mV VDDG which I thought was 'stable' after 120 runs of Linpack with <1gflops and 2 hours of VT3 with <0.05*10^09 variance, but apparently not because after a couple days I heard audio crackle in CS2 with Youtube.
I went back and tested combinations of 1.15-1.25VSOC, 0.95-1.15VDDP, and 900-1000VDDG and settled on the following which made my game feel the smoothest its ever been (until it isn't haha).
Does everything look good here? I'm not sure how to analyse Prime95 large FFTs but should there be variations in tests completed like that? And for Gflops / VT3 bits per second, is there a specific number you should aim for? Like my peak was 7.8*10^09 for 30 seconds run but it dropped a lot for next iterations, however, my most stable was around 7.6*10^09. Thanks!
Timings and everything has already passed 8 hours TM5 Absolut, 2 hours Extreme and 1usmus, along with 10,000% Karhu. Furmark mixed in-between different tests too. Linpack, VT3, and Prime95 ran in safe mode.