I am experiencing crashes everyday.
They are either VIDEO_TDR_ERROR in amdkmdag.sys, or THREAD_STUCK_INDEVICE_DRIVER in both dxgkrnl.sys and amdkmdag.sys, the later one happens simultaniously.
I am not gaming, it happens when it is iddling with only desktop, browsing, emailing etc.
DISM and sfc reports no violations or errors, checkdisk reports no errors
MSI MEG Z390 ACE
Corsair DDR4-2132 (CMK16GX4M2K4000C19) 4 modules , in total 32GB
Graphics Radon RX590 8GB
OS-Disk WD Black SN750 NVMe SSD M2
OS W10 20H2 19042.844
THis seesm to happen even when I run att BIOS default, i.e no overclocking, no XMP etc.
Windows and all devicedrivers, BIOS etc is up to date with latest available versions, (No beta-versions though)
my advice.. check if your cpu coolers metal bracket is touching the round condensators in the side of the motherboard or your m.2 nvme harddisk in the top slot.. If your cpu metal mounting system is touching anything that could be the reason for instability. ( i had this issue with my castle 365 v2 AIO water cooling system am4 mounting system.
Also maybe try another cpu cooler if you have one.
you can also try to take out your ram, cpu and gpu and insert it again.
Also check that the back of the motherboard is not touching the metal of your computer case or you dropped metal anywhere that can make an electric connection..
ALso try to remove all cables and insert them again.
And use two power cables from your power supply to your motherboard, and use two cpu power cables from power supply to top of the corner of your motherboard ( 4 pin and 8 pin )
ps. if you used metal liqued cpu paste then there is a risk you damaged your hardware.. never use that crap in a pc, just use ordinary arctic silver or other products.
Its a liquid cooler, so there is nothing that can interfere, as far as I can see, but will re-check.
Just ran a 10 hour memtest, with no errors.
Will check If I use two cables, beleive I do though.
Temperatures are ok, when it crashes, CPU and GPU both well under 50 degrees C (some 40+ or so)