The log file consists of umpteen pages of stuff we have to scroll through to in order to extract the info we need. When you open HWinfo, there is a button "Summary", it opens a summary of the installed hardware that you can take a screenshot of. But I have a feeling that I'd also have to explain how to take a screenshot, so let's check the stuff you uploaded before...
AMD Ryzen 5 5600G
MSI B550 GAMING GEN3
BIOS P.A0 09/03/2025
4x 16 GB F4-3200C16-16GIS DDR4-3200, running at DDR4-2666
Logitech USB keyboard, Logitech wireless mouse
Iiyama PL2282H monitor from 2013
KINGSTON SA400 960 GB SATA SSD as boot drive
TOSHIBA DT01ACA200 2 TB SATA HDD
TSSTcorp CDDVDW SH-224FB
We still need to know the PSU model. BIOS is updated as you said, good, although I would recommend the latest beta version, it brings an updated fTPM.
Now, a few things caught my attention in that log file. First, the RAM,
four modules always run worse than two. Granted, it's not running at XMP speed at the moment, but it might be worth an experiment with only two modules installed in slots A2 and B2, which immediately halves the stress on the memory system, to see if it has any involvement.
The monitor is also quite old, but the problem clearly comes from the PC if you also get reboots and part of the screen going black (as opposed to only full-screen black).
The SSD, that's one of the worst SATA SSDs out there, prone to failure, also see
here. Glancing at the SMART data, something also doesn't seem quite right on that end, high CRC error rate and what looks to be a bunch of bad sectors. This is a bad sign and a promising lead. I'd like you to use
CrystalDiskInfo to read out the SSD's SMART data properly again. In the program's menu under "Function" -> Advanced Feature -> Raw Values, select "10 [DEC]" to have human-readable values, and press CTRL-S to save a screenshot. Set to English as well.