PC crashes randomly : 'WHEA UNCORRECTABLE ERROR'

[InternalError]

Hi everyone,

First post here, glad to join the community

As of late, my PC crashes randomly and gives a BSOD, it is the first time that it happens in years of usage. Here is my configuration:

- CPU: AMD Ryzen 5 3600

- GPU: RTX 2070 SUPER

- Motherboard: MSI B450i GAMING PLUS

- RAM: Ballistix Sport LT 16GB (8GBx2 - DDR4 3200MT/s, CL16

I had this computer for 5 years and it worked correctly until now, so I am not sure of why it starts to fail .

I am currently running a dual boot with Windows 10 and Manjaro and my computer fails in both OSes.

The behaviours that I could observe are:

- On Windows, it fails directly and shows a BSOD. I am usually able to run for 10+ minutes. But when I am watching a video, it seems to fail quicker

- On manjaro, after few minutes of usage, the computer starts to fail. I still see the wallpaper, windows are still present, but I cannot interact with them (except moving them).

If I remember correctly it started after an update on Manjaro that failed and I had to hard reboot, but I am not sure.

 

On Windows I tried some steps on this guide (in french: https://recoverit.wondershare.fr/windows-10/fix-whea-uncorrectable-error-in-windows-10.html )

sfc ,  mdsched, chkdsk seems to show no issues.

So I suspect it might come from the GPU ? As it happens faster when I watch videos ?

Do you have any ideas of what I should check ?

[CommanderAlex]

Welcome to the forums!!!

 

Clear CMOS and run the RAM at default speeds and not XMP/DOCP. Check back if it's still crashing. 

[aisle9]

Are you overclocking your CPU or RAM?

[lafrente]

Can't be 5 years. That CPU was out mid 2019.

 

Most likely it's the CPU.

 

Try static OC'ing (It's called OC but not really) to 3.8 to 4.0 and set the voltage equal or lower than 1.2v in BIOS, find a stable combination and see if crashes persist.

[Applefreak]

31 minutes ago, InternalError said:

On manjaro, after few minutes of usage, the computer starts to fail. I still see the wallpaper, windows are still present, but I cannot interact with them (except moving them)

Are both OSes on the same storage device? 

[InternalError]

I did the CMOS Clear, I will keep you updated on if I see the issue

27 minutes ago, lafrente said:

Can't be 5 years. That CPU was out mid 2019.

 

Most likely it's the CPU.

 

Try static OC'ing (It's called OC but not really) to 3.8 to 4.0 and set the voltage equal or lower than 1.2v in BIOS, find a stable combination and see if crashes persist.

yeah you are right about the dates, my bad. But it has been a long time.

I will also try static OC.

 

29 minutes ago, aisle9 said:

Are you overclocking your CPU or RAM?

No overclock on my PC. My case is a SFF (Silverstone Sugo SG13), I prefered not to OC the computer.

 

21 minutes ago, Applefreak said:

Are both OSes on the same storage device? 

Yes, on a 1To SSD M.2 Samsung 980, I previously had a 250Go To SSD (with only windows), I did the update last year

[Applefreak]

1 minute ago, InternalError said:

I did the CMOS Clear, I will keep you updated on if I see the issue

yeah you are right about the dates, my bad. But it has been a long time.

 

No overclock on my PC. My case is a SFF (Silverstone Sugo SG13), I prefered not to OC the computer. I will also try static OC.

 

Yes, on a 1To SSD M.2 Samsung 980, I previously had a 250Go To SSD (with only windows), I did the update last year

Partitioning an SSD is not the best idea, as there are no real spaces on a HDD to allocate. Data of the inactive one will have to be moved around a lot to make room for bits of the active one. That causes problems. Do you still have the old SSD or maybe a live linux you can try to run so that you can rule out your SSD? The error is hardware related. A bad CPU would not boot or not properly anyways. So it could be bad storage, bad memory, bad motherboard (not likely but possible) or a bad PSU. GPU and PSU errors would occur when being taxed higher than average, so when launching a game for example. If both OSes crash after a certain amount of time, I suspect either storage or memory to be at fault here as these two are the only ones being taxed here with tasks in the background. Both windows and linux will write to the SSD (temp files, browser cache etc.) and also load up ram once you have logged on. To test out the GPU, install and open furmark. It will immediately tax your GPU. If it does not crash or freeze instantly, your GPU is most likely fine, as is your PSU as it would trip and shut down the system.

[InternalError]

23 hours ago, CommanderAlex said:

Welcome to the forums!!!

 

Clear CMOS and run the RAM at default speeds and not XMP/DOCP. Check back if it's still crashing. 

So I tried using my computer today after clearing the CMOS (yesterday). For now everything seems to work fine. Cases where it previously failed (watch a video for 10+min) are now working.

Thank you for your help guys !

[CommanderAlex]

4 minutes ago, InternalError said:

So I tried using my computer today after clearing the CMOS (yesterday). For now everything seems to work fine. Cases where it previously failed (watch a video for 10+min) are now working.

Thank you for your help guys !

No problem!!! If the issue comes back, we're here to help. 

[InternalError]

Hi again,

So this issue came back lately .

On 9/11/2023 at 12:14 AM, Applefreak said:

Partitioning an SSD is not the best idea, as there are no real spaces on a HDD to allocate. Data of the inactive one will have to be moved around a lot to make room for bits of the active one. That causes problems. Do you still have the old SSD or maybe a live linux you can try to run so that you can rule out your SSD? The error is hardware related. A bad CPU would not boot or not properly anyways. So it could be bad storage, bad memory, bad motherboard (not likely but possible) or a bad PSU. GPU and PSU errors would occur when being taxed higher than average, so when launching a game for example. If both OSes crash after a certain amount of time, I suspect either storage or memory to be at fault here as these two are the only ones being taxed here with tasks in the background. Both windows and linux will write to the SSD (temp files, browser cache etc.) and also load up ram once you have logged on. To test out the GPU, install and open furmark. It will immediately tax your GPU. If it does not crash or freeze instantly, your GPU is most likely fine, as is your PSU as it would trip and shut down the system.

I tested using furmark, and it ran for 10+min with the GPU stress test for 3840x2160 resolution, on the Task manager on windows GPU usage is around 70%. So GPU and PSU should be ok ?

I think that the RAM is ok, as the mdsched tool showed no issue.

So I guess the culprit is most likely the SSD ?

I don't have my previous SSD, it is not in my appartment (I will get it back when I can). So I cannot test it for now.

 

Other note: I decided to stop doing dual boot and installed Windows 11 to give it a try. Issue is still present and happened very quickly.

[Bjoolz]

Go to C:\Windows\Minidump and check if you have any minidump files. If you do, go back to the Windows folder and copy the Minidump folder itself to the Downloads folder (You can use the desktop if you don't have OneDrive syncing files). Zip the copied folder and attach it to a post. Please follow the instructions to the letter as Windows doesn't like you messing with files in this location.

[InternalError]

41 minutes ago, Bjoolz said:

Go to C:\Windows\Minidump and check if you have any minidump files. If you do, go back to the Windows folder and copy the Minidump folder itself to the Downloads folder (You can use the desktop if you don't have OneDrive syncing files). Zip the copied folder and attach it to a post. Please follow the instructions to the letter as Windows doesn't like you messing with files in this location.

It seems that I do not have a Minidump folder

[lafrente]

1 hour ago, InternalError said:

Hi again,

So this issue came back lately .

I tested using furmark, and it ran for 10+min with the GPU stress test for 3840x2160 resolution, on the Task manager on windows GPU usage is around 70%. So GPU and PSU should be ok ?

I think that the RAM is ok, as the mdsched tool showed no issue.

So I guess the culprit is most likely the SSD ?

I don't have my previous SSD, it is not in my appartment (I will get it back when I can). So I cannot test it for now.

 

Other note: I decided to stop doing dual boot and installed Windows 11 to give it a try. Issue is still present and happened very quickly.

My bet is still on the CPU. And about that quote which says "Bad CPU wouldn't boot" that couldn't be far from the truth. Bad CPU can boot perfectly, run perfectly until it doesn't and either cause shutdowns or random restarts.

 

Did you actually try static OC?

[InternalError]

51 minutes ago, lafrente said:

My bet is still on the CPU. And about that quote which says "Bad CPU wouldn't boot" that couldn't be far from the truth. Bad CPU can boot perfectly, run perfectly until it doesn't and either cause shutdowns or random restarts.

 

Did you actually try static OC?

I did not try static OC, I will do it

 

On 9/10/2023 at 11:35 PM, lafrente said:

Can't be 5 years. That CPU was out mid 2019.

 

Most likely it's the CPU.

 

Try static OC'ing (It's called OC but not really) to 3.8 to 4.0 and set the voltage equal or lower than 1.2v in BIOS, find a stable combination and see if crashes persist.

For the 3.8 to 4.0 value: you mean the Frequency in GHz ?

 

Edit: I just set the CPU voltqge to 1.2V and it seems to not boot (everything seems on but nothing is on the screen, not even the BIOS )

Edit 2: I was able to reboot the PC by resetting the default setting through a CMOS Clear, and then I tried 1.2v voltage and 3.8GHz frequaency for base clock, same result it does not boot.

Another detail: I never overclocked my computer so I am not very familiar with the pitfalls of overclocking. If you have a great guide that you could recommend, I am interested.

[Bjoolz]

14 hours ago, InternalError said:

It seems that I do not have a Minidump folder

As you are not getting dump you can make the BSOD screen display more info.  

 

To make the BSOD screen display the additional info on the BSOD screen we need to add a field to the registry. If you are not comfortable editing the registry then do not do this step. Navigate to

HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\CrashControl

right click on the empty area on the right section and select New → DWORD value with the name "DisplayParameters". Right click on it, modify and set the value data to 1 (Does not matter if you use Hexadecimal or Decimal). It should look like this once done. Reboot to apply the registry change."

 

The next time you BSOD, you should have these extra numbers in the top left corner. I want a picture of this. If it's automatically restarting and you need more time at the BSOD screen for the picture, look at this guide for changing dump type and at this screen uncheck automatically restart. Just use the power button to restart. 

[InternalError]

On 10/8/2023 at 9:34 AM, Bjoolz said:

As you are not getting dump you can make the BSOD screen display more info.  

 

To make the BSOD screen display the additional info on the BSOD screen we need to add a field to the registry. If you are not comfortable editing the registry then do not do this step. Navigate to

HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\CrashControl

right click on the empty area on the right section and select New → DWORD value with the name "DisplayParameters". Right click on it, modify and set the value data to 1 (Does not matter if you use Hexadecimal or Decimal). It should look like this once done. Reboot to apply the registry change."

 

The next time you BSOD, you should have these extra numbers in the top left corner. I want a picture of this. If it's automatically restarting and you need more time at the BSOD screen for the picture, look at this guide for changing dump type and at this screen uncheck automatically restart. Just use the power button to restart. 

I am currently trying this, however since enabling "DisplayParameter". I was not able to reproduce the crash (an update might have been done during the reboot). I will keep you updated