Question Weird computer crash when playing (most) games

[ilyxas]

This is a copy paste of a reddit post I also made


Specs
Windows 11 (this was happening on windows 10 as well)
Nvidia rtx 2080
Ryzen 5800x (not over clocked)
Boot drive - Samsung 970 evo
So a few months ago I was having my audio cut out in games, some sounds would be quiet some sounds would be impossible to hear for instance in valorant I could hear footsteps but sometimes I couldn't hear guns shooting and it would sound muffled as well, then after a few minutes the game would become unresponsive but If I looked at my left monitor I could see everything running fine I could see Spotify playing but I couldn't hear it I could see discord messages but couldn't hear or speak in it my background would still be animated etc but when I clicked into any of those windows it would completely crash and become unresponsive so I had to restart my pc manually and sometimes when I would load into windows it would make me re sign into spotify and discord and stuff.
now to fix that I reinstalled windows fresh on a new drive and took out the old drive and it happened again so I installed windows 11 and fresh reset again (every drive this time including drives just meant for games or pictures) and its still happening all my drivers are up to date I've toggled a million settings on and off and no fix I have safe boot off I've swapped my ram for my girlfriends ram (which was an upgrade) I took every other drive out I've reinstalled obviously.
forgot to mention I also monitor in task manager and in hwmonitor everything going on, my cpu is around 20 percent, my ram is around 20 percent, my gpu is around like 30 percent, my drives don't spike in usage or anything everything seems to be completely okay with the sysem but its gotten so bad I can't even load into a game for more than 2 minutes without crashing now.
honestly any help would be incredible I don't really have the money to go to any tech repair place because I just moved and had to take my cats to the vet so any help is a lot of help at this point ;-;

 

[DavidM012]

What's your:

Cpu Cooler:
Mobo : Make and model
PC case :
Any optional fans:

Power supply unit: (be particularly specific on make, model and age).

 

[ilyxas]

What's your:

Cpu Cooler:
Mobo : Make and model
PC case :
Any optional fans:

Power supply unit: (be particularly specific on make, model and age).

Cooler : corsair h100i
Case : Lian Li mesh C3
Fans : 3 corsair ll120 fans in the front 2 in the top on the rad and one at the back the front 3 are intake the back two and top two are exhaust
Mobo : Asus x570-Pro
Psu : EVGA 850 Supernova T2 220-T2-08050-X1 80+Titanium a couple years old

 

[DavidM012]

That all looks solid. You switched the dimms and the drives and reinstalled windows.

So maybe there's something with the n.v.me. drive port, or mobo. Samsung Magician drive diagnostics all good?

Do you have any usb peripherals attached, extenders or things like that? Anything with wires that could be worn or not properly seated or connected?


How about... patch

https://www.tomshardware.com/news/nvidia-issues-fix-for-windows-11-22h2-lag-and-stuttering-bugs

 

[ilyxas]

That all looks solid. You switched the dimms and the drives and reinstalled windows.

So maybe there's something with the n.v.me. drive port, or mobo. Samsung Magician drive diagnostics all good?

Do you have any usb peripherals attached, extenders or things like that? Anything with wires that could be worn or not properly seated or connected?


How about... patch

https://www.tomshardware.com/news/nvidia-issues-fix-for-windows-11-22h2-lag-and-stuttering-bugs

I have my keyboard and mouse and headphone wireless adapters plugged in, I downloaded and tried to run a diagnostic scan on my 970evo and it said it didn't support it so Idk about that but the performance benchmark came back pretty good its not hot or anything either

 

[DavidM012]

Ok if doesn't support the diagnostic you simply can't run that. Some do and some don't. That's normal. Can I assume that your bios is fully up to date from the mobo manufacturer's site?

So maybe just the nvidia patch. Although your symptoms started before the windows update.

Otherwise I'm stumped. You have a good combination of hardware so there's nothing obvious.

You can monitor the cpu usage in task manager during a prime 95 stress test to see if the cpu dips under 100% . Use version 26.6 first of prime 95 since that's non avx. Later versions also add avx instructions which some say is an 'unrealistic workload'. The 5800x supports avx2 but stressing it with ax2 will also make it much hotter. You have good cooling but you haven't believed a stress test 'ti you've seen it. Really does max out your cpu. Also anything that's less that optimal could potentially fry. A stress test will be fine on a healthy system with excellent cooling.

So take some time to get used to it. If the system's stable a stress test should be able to go for a couple of hours without crashing. If you aren't too comfortable with that then just run it for 5-10 minutes or stop if you see any cpu usage dips in task manager, whatever happens first.

There's a tutorial in the overclocking section of the forum on how to run the stress test. You aren't overclocking but the stress test is the same.

There's a reddit that talks a bit about that too.

If the cpu is dipping in task manager it's probably something to do with the mobo vrms. Don't let the cpu temperature exceed 95c or if it's getting close to that stop the test. 85c is probably a better place to stop. So you have to monitor temps. while stressing.

Going beyond the obvious well only a 2 yr old quality psu, you'd expect would hold up. Could try another different psu, or different mobo if you're getting cpu dips. Thus the problem it's not easy to have spares lying about.

The only other thing I can think of is memory compatibility. Maybe your dimms aren't exactly qvl for it. I mean exactly specified on the mobo manufacturer's memory support list. However, your system was running fine until a few months ago so, I'd just check to be sure.

So before digging into the 'stressing' install the nvidia beta patch from the aforementioned article and if that doesn't help, well I'd be thinking probably hardware with maybe mobo or psu suspect.

 

[ilyxas]

Ok if doesn't support the diagnostic you simply can't run that. Some do and some don't. That's normal. Can I assume that your bios is fully up to date from the mobo manufacturer's site?

So maybe just the nvidia patch. Although your symptoms started before the windows update.

Otherwise I'm stumped. You have a good combination of hardware so there's nothing obvious.

You can monitor the cpu usage in task manager during a prime 95 stress test to see if the cpu dips under 100% . Use version 26.6 first of prime 95 since that's non avx. Later versions also add avx instructions which some say is an 'unrealistic workload'. The 5800x supports avx2 but stressing it with ax2 will also make it much hotter. You have good cooling but you haven't believed a stress test 'ti you've seen it. Really does max out your cpu. Also anything that's less that optimal could potentially fry. A stress test will be fine on a healthy system with excellent cooling.

So take some time to get used to it. If the system's stable a stress test should be able to go for a couple of hours without crashing. If you aren't too comfortable with that then just run it for 5-10 minutes or stop if you see any cpu usage dips in task manager, whatever happens first.

There's a tutorial in the overclocking section of the forum on how to run the stress test. You aren't overclocking but the stress test is the same.

There's a reddit that talks a bit about that too.

If the cpu is dipping in task manager it's probably something to do with the mobo vrms. Don't let the cpu temperature exceed 95c or if it's getting close to that stop the test. 85c is probably a better place to stop. So you have to monitor temps. while stressing.

Going beyond the obvious well only a 2 yr old quality psu, you'd expect would hold up. Could try another different psu, or different mobo if you're getting cpu dips. Thus the problem it's not easy to have spares lying about.

The only other thing I can think of is memory compatibility. Maybe your dimms aren't exactly qvl for it. I mean exactly specified on the mobo manufacturer's memory support list. However, your system was running fine until a few months ago so, I'd just check to be sure.

So before digging into the 'stressing' install the nvidia beta patch from the aforementioned article and if that doesn't help, well I'd be thinking probably hardware with maybe mobo or psu suspect.

Okay so after trying all this stuff along with a lot more stuff my girlfriend randomly just said its the ssd (she knows nothing about computers) so I just took it out installed windows to a different ssd and all is good now ;-;

 

[DavidM012]

Well it is a process of elimination so at least you know your cpu and mobo were ok then you could review some earlier steps. That wasn't easy.

If samsung magician says the drive is good, and the warranty is 5 years how can you know otherwise? So you basically found out everything else was good so suspected the drive.

Needle in a haystack, got there in the end.