Question PC stuck in a (finite) boot loop / randomly crashes when testing for errors after

[Vagnard]

Hello, I think I have my problem narrowed down at least but I would like a few opinions before I go and buy new parts if you would be so kind.

What happend:
Got 2 application crashes when I was rendering a video edit yesterday. Didnt think much of it.
Then I got a full system crash when I was watching a stream and had an Android emulator open.
I needed to manually reboot only for the system to crash again upon reaching the Win10 login screen.
Manually reboot again and now the loops begin to occur. Power light goes on, CPU fan revs up, power light goes out, repeat.
After a random number of times the system boots "normally". If this goes on for too long I kill the power manually and try again later.

What I tested so far:
no visibly blown or bulging capacitators on the board
full system virus scan -> no issues
CPU temp at high load -> normal, or lets say normal for Haswells
GPU temp on medium load -> normal
!! Intel diagnostics tool -> full crash upon reaching float point test
!! 7 instances of memtest -> full crash at about 5% each
!! trying to watch troubleshooting video in Chrome -> full crash

What I think I could be:
RAM -> 1st choice. The memtest crash + the random app/system crashes when using a lot of it all seem to point here
CPU -> unlikely, but not impossible. The Intel tool threw me off for sure. No obvious temp issues though
PSU -> possible. I did the fan/wall outlet test and it ran, also it apparently can still power the system to an extent but I dont have a Multi tool to properly test the pins to see if the board gets enough juice.
Board -> possible. No visible issues but I dont really know how to test for anything further here.

My system is from 2014 so there certainly is room for a critical hardware failure based on age alone
i7 4470K not overclocked
16 GB DDR3 dual @ 1333MGHz
MSI MS 7821 board
Nvidia GTX 770 4GB
bequiet! 550W

If you diagree with my assesment, have another idea or think I missed something or should absolutely test something else please let me know.
The plan (for now) is to take apart the system tonight and test my 2 RAM sticks solo then go from there.

 

[Lutfij]

Welcome to the forums, newcomer!

How old is the PSU? If it's from 2014, You forgot to mention the model of your MSI motherboard, gaming can't denote what model the board is specifically. Are you on the latest BIOS version for your motherboard? You also will need to mention the make and model of your GPU.

 

[Vagnard]

Oh sorry. Thanks for the welcome.

The PSU is from 2014 also, bought all part together back then
The board is MSI MS-7821, I'll check for a BIOS update when I get home again
The GPU is a Nvidia GTX 770 4GB.

 

[Vagnard]

Ok, I did a BIOS update and promptly experienced the boot loop again. Luckily the update still went through as far as I can tell.

I did another Intel diagnostics run with Core temp open and this time it didnt crash, but it actually failed the Prime number test and the GPU Stress-W (whatever that is). Yesterday, before the update, it passed the prime number test, but then crashed during float point. No abnormalities regarding core temperature btw.

I'll be doing another mem test run now and maybe play a bit of musical chairs with my RAM but its already pretty late here unfortunately.

 

[Vagnard]

Well, I'll be dammed. Memtest ran all the way through, no errors, no crashes.
So it might not be the RAM after all.
I guess I'll do a bit of CPU stress testing now with prime95, see if that gives me a crash again.
Bootloop is still there unfortunately.