Ya I saw you have 1800w, I was just correcting RYXES.

If a rig is unstable like you say, it may be risers or motherboard. Look at logs see if it is one GPU in particular and try changing riser. Can also unhook all but one, and incrementally test. In Windows can use HwInfo64 to check for mem errors and other things. I had a Biostar TB250 mobo would throw lots of PCIE errors, resulting in about 20 reboots per day. Replaced mobo, rock solid. I don't know if there is a Linux tool. My experience with Linux has been figuring out the desktop is a PITA, and VMs/Server for webdev work.