computer keeps restarting after gaming sessions

M

MartianPotatoSoup

Member
My computer keeps restarting.
This has been primarily after I finish a gaming session, but occasionally during.
So usually it happens maybe 1 - 2 or 3 minutes after I have gone back to windows. I just got xcom 2 the other day (steam sale) and it restarted twice at the exact same point ingame on the first map.

Im guessing it has something to do with the card (2080 Ti - less than 2 years old) as this seems to only happen after 3d games (even things like project winter, but weirdly enough not after gta online..). But surely if the card was broken it just wouldn't run, or at least crash back to windows rather than restart the whole system?

The most strenuous game Ive played on this machine is a modded skyrim ( living skyrim, via wabbajack). Card is updated to the one before most recent drivers. This has been affecting me for about 6 months, but I feel is happening more often now.

I have checked temps of the card (hitting about 83 on skyrim), and fans go off every minute or so. I have had hours long sessions with skyrim, no issue. cpu and rest of computer temps all seem normal, fans all normal.

So is the card broken, or could there be some other thing I could investigate?

Am waiting on response from pc specialist tech support a few days, but this seems quicker if there is anything i can do meantime.

Case
FRACTAL DEFINE R6 BLACK QUIET MID-TOWER CASE
Overclocked CPU
Overclocked Intel® Core™ i9-9940X 14 Core (3.30GHz @ up to 4.4GHz)
Motherboard
ASUS® PRIME X299-A: ATX, USB 3.1, SATA 6 GB/s, RGB Ready
Memory (RAM)
64GB Corsair VENGEANCE DDR4 2666MHz (4 x 16GB)
Graphics Card
11GB NVIDIA GEFORCE RTX 2080 Ti - HDMI, 3x DP GeForce - RTX VR Ready!
1st Storage Drive
240GB ADATA SU630 2.5" SSD, SATA 6 Gb (520MB/R, 450MB/W)
1st Storage Drive
2TB SEAGATE BARRACUDA 2.5" SSD, (up to 560MB/sR | 540MB/sW)
1st Storage Drive
6TB SEAGATE BARRACUDA PRO 3.5", 7200 RPM 256MB CACHE
Power Supply
CORSAIR 850W RMx SERIES™ MODULAR 80 PLUS® GOLD, ULTRA QUIET
Power Cable
1 x 1 Metre UK Power Cable (Kettle Lead)
Processor Cooling
Corsair H100x Hydro Cooler w/ PCS Ultra Quiet Fans
Thermal Paste
COOLER MASTER MASTERGEL MAKER THERMAL COMPOUND
Sound Card
ONBOARD 8 CHANNEL (7.1) HIGH DEF AUDIO (AS STANDARD) (External ASIO soundcard)
USB/Thunderbolt Options
2 PORT (2 x TYPE A) USB 3.1 PCI-E CARD + STANDARD USB PORTS
Firewire
2 Port IEEE 1394a Firewire PCI-E Card (2 x 6 pin)
Operating System
Windows 10 Professional 64 Bit
Operating System Language
United Kingdom - English Language
Windows Recovery Media
Windows 10 Multi-Language Recovery Image - Unlimited Downloads from Online Account
Anti-Virus
Malware Bytes/win defender
Monitor
MSI 31.5" Optix MAG322CQRV - 2560 x 1440, 1MS, 144Hz
Keyboard & Mouse
Viper V65, Razor Deathadder Elite
 
SpyderTracks

SpyderTracks

We love you Ukraine
More likely driver issues, usually if a GPU fails, you'll get artifacting / off colours before the crash. It could be PSU related also given the symptoms.

Can you upload you dump files?

The memory dump file is typically located in %SystemRoot%\MEMORY. DMP. The system root is typically C:\Windows If you've configured the system for a minidump, the default location folder is %SystemRoot%\Minidump.

OOoooo, just noticed the cooler on that chip.... have you monitored CPU temps in game? I wouldn't be happy with that cooler personally.
 
M

MartianPotatoSoup

Member
Only Memory dump is 1.75Gb

I had setup the minidumps before, but I think the recent windows update has reset that, so no minidumps.

Whocrashed says:
On Sat 06/02/2021 13:29:23 your computer crashed or a problem was reported
crash dump file: C:\WINDOWS\LiveKernelReports\PoW32kWatchdog-20210206-1329.dmp
This was probably caused by the following module: ntkrnlmp.exe (nt!IoUnregisterShutdownNotification+0xF84)
Bugcheck code: 0x1A1 (0xFFFFB98FE8A4E080, 0x0, 0x0, 0x0)
Error: CUSTOM_ERROR

The crash took place in the Windows kernel. Possibly this problem is caused by another driver that cannot be identified at this time.

Usually after such resets there is no new information in whocrashed.

As mentioned in original post I monitored the temps when playing skyrim (using open hardware monitor - I have an excel from this session). GPU was hitting 83; CPU and HHD were all around 40 -50 iirc.

The cooler on the chip - do you mean the thermal paste? I guess it was that kind of advice I was expecting from PC Specialists when I put the order in. Not going to lie, this system, for what it is, has always felt badly optimized.
 
jamiephillips909

jamiephillips909

Godlike
MartianPotatoSoup said:
Only Memory dump is 1.75Gb

I had setup the minidumps before, but I think the recent windows update has reset that, so no minidumps.

Whocrashed says:
On Sat 06/02/2021 13:29:23 your computer crashed or a problem was reported
crash dump file: C:\WINDOWS\LiveKernelReports\PoW32kWatchdog-20210206-1329.dmp
This was probably caused by the following module: ntkrnlmp.exe (nt!IoUnregisterShutdownNotification+0xF84)
Bugcheck code: 0x1A1 (0xFFFFB98FE8A4E080, 0x0, 0x0, 0x0)
Error: CUSTOM_ERROR

The crash took place in the Windows kernel. Possibly this problem is caused by another driver that cannot be identified at this time.

Usually after such resets there is no new information in whocrashed.

As mentioned in original post I monitored the temps when playing skyrim (using open hardware monitor - I have an excel from this session). GPU was hitting 83; CPU and HHD were all around 40 -50 iirc.

The cooler on the chip - do you mean the thermal paste? I guess it was that kind of advice I was expecting from PC Specialists when I put the order in. Not going to lie, this system, for what it is, has always felt badly optimized.
Click to expand...
Processor Cooling
Corsair H100x Hydro Cooler w/ PCS Ultra Quiet Fans = cooler on chip :)
 
ubuysa

ubuysa

The BSOD Doctor
This is a most unusual BSOD. The stop code is a common one, SYSTEM_THREAD_EXCEPTION_NOT_HANDLED, which means that a system thread (as opposed to a user thread) generated an exception that the error handlers could not handle. What is unusual here is the exception code of 0x80000003, which indicates that a breakpoint has been reached in the code but the system was running in the normal (ie. /nodebug) mode and thus no debugger was connected to handle the breakpoint.

The stack trace of the active thread shows the nt!_C_specific_handler function being called, I assume (because I don't know) that this is related to some code written in C? Beta driver code perhaps? Shortly afterwards in the stack trace is the nt!KiBreakpointTrap call, this function is being called because a breakpoint was found in the code. Immediately after the nt!KiBreakpointTrap call we see the exception handler being called (because the breakpoint can't be handled).

In the list of driver calls in the active thread we get an Unable to load image \SystemRoot\system32\drivers\PrismSoundWDM_Win10_X64.sys, Win32 error 0n2 error and that Prism Sound driver is called many times in the thread's driver calls list.

I would suggest that you look at the Prism Sound drivers, I wonder whether you're running a beta version of a Prism driver? I cannot imagine how you're running code that still contains breakpoints? That's normally reserved for the developers....

I can also see that you're running Malwarebyte's Antimalware Pro, it might be worth checking that MBAM isn't affecting Prism Sound in some unexpected way too.

Later Edit: We have had reports of conflicts between the Nvidia audio driver and the standard Realtek audio driver. I wonder whether you have the Nvidia audio driver installed and it's also conflicting with Prism Sound? It might be worth removing the Nvidia driver with DDU and reinstalling it without the Nvidia audio driver...?
 
Last edited:
M

MartianPotatoSoup

Member
Interesting, PrismSoundWDM was referenced before a couple of times in Whocrashed (before the win update). I had recently updated the hardware firmware and drivers/control software in response to this (as far as I can tell, not beta, most recent update last year). I will check the other points you mentioned, but I guess first port of call will be to assign primary default audio to one of the internal soundcards, and leave the prism out of that, just for the audio software i use.

Thank you very much for the detailed breakdown. Will report back to how things have fared in due course.
 
M

MartianPotatoSoup

Member
Ok, so I removed all the nvidia drivers, then reinstalled (latest drivers) without audio. I noted that there still seemed to be a left over nvidia high quality audio driver that had no uninstall button, but I uninstalled in the device manager. I set default audio to the realtek. Next I uninstalled the Prism driver, and control panel. I also checked the \system32\drivers\ folder and noted that PrismSoundWDM_Win10_X64.sys, and another Prism related driver were still there, so I manually removed these too. I made sure that the system was writing a complete memory dump - seems it was already set to that. I just noticed that the Memory.dmp file I sent the other day hadn't been modified since 4/12/2020 (?)

Results were, I got a bit further in xcom than originally, but it crashed and I had to physically hit the reset button - no ctrl alt delete would work. Seemed mainly in cut scenes or in the base segment of the game (ie - not in the tactical shooting aliens part). I did about 20 minutes of skyrim, looked ok, but then then computer restarted after I finished and had exited skyrim... Minecraft seemed fine (about 20 min)

Tomorrow I will uninstall malwarebytes, see how that goes, but im not hopeful...

Don't know if this is pertinent, but I mentioned in my original post that the system has always felt under optimized. I guess the first point was that PC Specialist didn't seem to have overclocked the machine as I had ordered. They sent me a file I had to load into the bios later on after I pointed this out a few weeks later, but if they missed something like that, they could have easily not bothered to have configured the machine up properly. (Its barely OC'd as it is anyway....) Load times from switching the machine on to getting to the start screen can vary **greatly** 30 sec - 1:30 in the worst cases, but usually somewhere in between. The image on the start screen is supposed to change - I dont think mine has ever changed, and I even followed instructions to clear the image cache so that it would start rotating as windows 10 does..never worked... Finally, Malwarebytes is set to run a system scan every morning. And when it does the whole machine comes to a crawl. I used malware bytes on my old machine, and ok, it was win 7, but it never affected the performance to the degree it does on this machine. This is things like just opening firefox or word, and the system almost freezes (you can move mouse, but clicking does nothing).

The main thing I would like to rule out is hardware failure... Then final course of action is to nuke the whole computer and do a clean install...

I should also add, Everything on this machine is legit, nothing is pirated. I try and take care of which websites I visit.
 
Martinr36

Martinr36

MOST VALUED CONTRIBUTOR
use DDU to totally uninstall the display driver

Display Driver Uninstaller (DDU) – Best Graphics Driver Removal Tool

Display Driver Uninstaller or DDU is a free graphics card driver removal tool for Nvidia and AMD graphics cards. DDU can completely remove video drivers if you cannot remove it from the windows uninstaller. Download Display Driver Uninstaller (DDU) for Windows. DDU is an effective and...
graphicscardhub.com graphicscardhub.com
 
ubuysa

ubuysa

The BSOD Doctor
MartianPotatoSoup said:
Ok, so I removed all the nvidia drivers, then reinstalled (latest drivers) without audio. I noted that there still seemed to be a left over nvidia high quality audio driver that had no uninstall button, but I uninstalled in the device manager. I set default audio to the realtek. Next I uninstalled the Prism driver, and control panel. I also checked the \system32\drivers\ folder and noted that PrismSoundWDM_Win10_X64.sys, and another Prism related driver were still there, so I manually removed these too. I made sure that the system was writing a complete memory dump - seems it was already set to that. I just noticed that the Memory.dmp file I sent the other day hadn't been modified since 4/12/2020 (?)
Click to expand...
Ah, so that kernel dump may not be related to the problems you're having now?

Are the any minidumps related to the problem in the folder C:\Windows\Minidumps? If so upload those to the cloud with a link to them here.
MartianPotatoSoup said:
Results were, I got a bit further in xcom than originally, but it crashed and I had to physically hit the reset button - no ctrl alt delete would work. Seemed mainly in cut scenes or in the base segment of the game (ie - not in the tactical shooting aliens part). I did about 20 minutes of skyrim, looked ok, but then then computer restarted after I finished and had exited skyrim... Minecraft seemed fine (about 20 min)
Click to expand...
This sounds more graphics related than anything else. When you removed the Nvidia drivers did you use DDU? If not please download it and uninstall the Nvidia drivers again with DDU (the PC will reboot). Then reinstall the latest driver from the Nvidia website.
MartianPotatoSoup said:
Tomorrow I will uninstall malwarebytes, see how that goes, but im not hopeful...
Click to expand...
I don't think Malwarebytes is the problem, it's generally a very good tool. It is worth removing it as a test however. :)
MartianPotatoSoup said:
Don't know if this is pertinent, but I mentioned in my original post that the system has always felt under optimized. I guess the first point was that PC Specialist didn't seem to have overclocked the machine as I had ordered. They sent me a file I had to load into the bios later on after I pointed this out a few weeks later, but if they missed something like that, they could have easily not bothered to have configured the machine up properly. (Its barely OC'd as it is anyway....) Load times from switching the machine on to getting to the start screen can vary **greatly** 30 sec - 1:30 in the worst cases, but usually somewhere in between. The image on the start screen is supposed to change - I dont think mine has ever changed, and I even followed instructions to clear the image cache so that it would start rotating as windows 10 does..never worked... Finally, Malwarebytes is set to run a system scan every morning. And when it does the whole machine comes to a crawl. I used malware bytes on my old machine, and ok, it was win 7, but it never affected the performance to the degree it does on this machine. This is things like just opening firefox or word, and the system almost freezes (you can move mouse, but clicking does nothing).
Click to expand...
I would speak to PCS about the overclocking, though others with more experience of it than I have may be able to offer better advice.

The varying boot times is a worry. I would first stop MBAM doing any automatic scans so you know that's not impacting anything, then open Task Manager, click the Start-up tab and disable anything in there that you don't actually need at startup. You'll lose some functionality of course, but with the minimum startup items enabled see whether you still get variable boot times.

I would also open a command prompt or PowerShell session and run a chkdsk /f /x on each of your drives - this will check and repair the NTFS filesystem on them (problems there can cause all sorts of niggly issues). You will have to reboot after each drive has been checked.

An MBAM scan slowing things to a crawl does point to a drive issue. Is the slowdown only when scanning files on one drive or on all of them?
MartianPotatoSoup said:
The main thing I would like to rule out is hardware failure... Then final course of action is to nuke the whole computer and do a clean install...

I should also add, Everything on this machine is legit, nothing is pirated. I try and take care of which websites I visit.
Click to expand...
A fully clean install, from bootable media and deleting existing UEFI partitions is the only reliable way to eliminate software as the cause. When you do a reinstall allow Windows Update to install all drivers and then manually download and install the latest Nvidia driver.

One other thing it's worth trying is to remove two RAM sticks for a day or two, then swap to the other two RAM sticks. That might indicate whether these issues are RAM related.
 
M

MartianPotatoSoup

Member
You’ve disabled the trial on MBAM?
Click to expand...
No, I have the premium version. Is this a one or the other type situation? Defender doesn't do scans, and I havent really touched it. (I note in your other post, it is said defender is sufficent, and I am aware of this). I figuered my biggest threat could be potential drive bys from compromised legit websites, hence MalwareBytes.
 
M

MartianPotatoSoup

Member
nvidia driver uninstalled with DDU, latest driver installed, no audio.
chkdsk run on all drives, everything looks ok.
Problem persists.
Looks like its a reinstall :(
 
M

MartianPotatoSoup

Member
Xcom has been running fine. Tested/played a few other games that had also caused issues and so far no crash after exit. So, so far, so good. Startup seems much smoother.

Barely anything is installed, only the graphics driver and steam.

One thing I have found interesting is the GPU fan hasn't been going like crazy during gaming (everything tested so far is "relatively" old). Is it possible that something else was putting the GPU under additional stress when I was playing GPU intensive games? And then when the game released the GPU resource it overloaded itself causeing a restart (Pure speculation, I have no idea about these things)?

Thanks for the assistance. I will slowly load back all my programs (its going to take at least a week if not more... :( ), and will leave final report in due course.
 
You must log in or register to reply here.
Top