Originally Posted by RedBlueHue4
So I have swapped GPU positions, no change. I have also done a complete re-install of windows OS in case of something application wise causing the issue - still having this problem. Next I am going to take out the other GPU and test it individually. I am starting to believe the issue either lies in the CPU or the motherboard. I am quite upset at this point, when I test the GPU 1 by 1 do I need to also remove the power supply cables from the PSU unit itself? Ex.) unhook GPU #2 but leave the PSU cables plugged in to the PSU unit? (its a pain in the ass for me to take them out/put them in but I don't want to cause any damage)
Also this issue has been going on for some time, since the build honestly.
No need to pull out any of the gpus. Just unplug the power from the gpu that is not being tested. But, like i said, do not forget to plug the monitor cable to the right gpu, The one being tested. When you game or bench, check the temps, voltages, etc. using HWINFO. Disable all other apps. You can close Afterburner once ULPS is set. See example. Those are my gaming temp. Check out the gpu voltage at load and all other readings.
Back to Afterburner. In the graph, when polling rate is set to 1000, the GPU clocks should be a straight line, tiny dips during cut scenes. If it shows jitters, then there might be something wrong. Even during games, the gpu clocks when loaded should be straight. The usage will jump up and down but it should not be too low.
BTW, what Platinum PSU is that? You are not using AI suite, are you? How about Ryzen Master?
By Built-in Benchmark Queen:
My 2.5 years old delidded 7980xe @ 4600mhz all core with 49ns memorylatency (4000c16 twekaed) is faster than 3950x max overclocked on water in every scenario. Pretty good for a ancient cpu
In gaming it has no chance vs my new 10900k @ 5400mhz and 4600c16 tweaked memory. ~35ns memorylatency.