Why does the system hang tightly and stress-ng?

Question

linux

Good day to all!
Guys, from time to time the computer freezes tightly, only a reboot is required. The load is small, the pieces are spinning. 3-4 virtual machines (virtualbox). Allocate them RAM no more than 4 GB.
System: debian 9.8.0 amd64 (stretch)
Hardware: processor - AMD Athlon(tm) II X2 250 Processor, video card - GeForce 7950 GT 512mb, power supply - exegate 500W, hard drive - Western Digital Caviar Green 1TB WD10EARS 1TB SATA , motherboard - ASUS M4A88T-M, RAM - 8GB.
The logs are useless.
I ran the RAM through Memtest86 once, no errors were found.
At first, everything sinned on the hard drive, usually they have such signs, but I ran it through the HDD Regenerator (started from the boot disk) and badblocks - no errors were found, but Victoria HDD (connected it to windows second) - found one damaged (red) bad- block and a couple of poorly readable ones (orange). I ran Victoria twice, the program fixed everything, but the problem remained.
Yesterday I added another cooler, I thought it might be warming up, now there is a refrigerator inside, the system hasn’t hung yet, but there was a spontaneous reboot once.
I did stress testing of the entire system through stress-ng, but as a result I can’t understand whether this program revealed bugs or not. Don't look?
Stress testing!
Thank you!

Answer 1

Such problems can be caused by a faulty or unsuitable PSU. When the load is high, the PSU starts to produce voltages that are insufficient to power the PC components, which leads to various kinds of failures. And, worst of all, the number of watts on the PSU label does not always indicate its quality or ability to power a powerful PC. Over the years, the problem has only gotten worse.
According to the stress-ng log, unfortunately, I can’t suggest anything.