Question BSOD - MEMORY_MANAGEMENT and PAGE_FAULT_IN_NONPAGED_AREA

[ubuysa]

As above. The way to confirm this is an Nvidia issue is to remove the Nvidia card and see whether it's stable on the Intel iGPU.

 

[sati-edimax]

I've removed Nvidia GPU yesterday, at first boot I've also changed in BIOS the amount of memory assigned to Intel GPU (from 64M -> 512M , in Windows it seems that gives me ~2.1GB of VRAM in Task Manager > Performance > Intel GPU).

Windows booted at first attempt without issues, no freezes/no forced reboot required like in the last 2 days with Nvidia and for the whole day I didn't have any "display driver stopped working" errors/warning in Event Viewer. I'm going to test how it looks for few days with only Intel GPU.

 

[ubuysa]

That's very wise.

If/when you decide that it's OK without the Nvidia card, I would suggest downloading the three most recent driver versions for the Nvidia card, and DDU. Install the Nvidia card, use DDU to uninstall the current driver, and then try each of the last three driver versions (using DDU between each install) and see whether it's stable on any of those drivers. If it's not then you're almost certainly looking ar a hardware problem with the Nvidia card

 

[sati-edimax]

A quick update:

Almost a week running only on Intel GPU and so far no issues - Windows boots without issues, no freezes, no lags, no display errors, no driver display errors/warning in Event Viewer, no "display driver stopped working" errors. Also no BSODs. I'll run this for few days more and then will do what @ubuysa said in #28

Cheers

 

[sati-edimax]

I think I'll stop writing "quick updates" as its another time that when I wrote one something else breaks (after the previous one the GPU started to give problems)

I was doing the usual stuff (wanting to still test the Intel GPU for few more days) and out of the blue the PC just shutdown. After about 2 hours of running on. Like total power shutdown as even the fans stopped spinning and after 2-3 seconds the PC booted itself on its own (I didn't press the power button at all). No BSOD, no visible freeze, no errors before the shutdown. No MOBO alerts on temp (and those were working when I was doing tests in OCCT).

After booting into Windows, I tried to check Event Viewer/Reliability Monitor to see if something was there but didn't manage to open them (both were at loading data phase) as the PC did the same thing - power off, 2-3 seconds and power on again. I quickly turned it off at BIOS splash.

Today it booted into both BIOS and Linux Live from USB (not running Windows from the SSD drives to not damage them with random power loss) without any issues. I'm now in Linux Live and running "System Stability Tester" to just have some CPU usage (the usage goes up to 50%, depending on the test setup), the CPU temps are 28-35 min, 73-82 max (at least that's what "Psenor - Temperature Monitor" shows). I might also plug the Nvidia GPU to have some additional "power hungry" hardware for more power usage. I also re-plug the power cables and verified if other cables aren't "stretching" the power cable).

I've checked and my BIOS setting on "AC BACK" is set to Always Off (System stays off upon power return). No idea why it didn't stay off and booted on it's own. I thought that maybe my UPS was at fault (the UPS is less than one year old, Ever Sinline 1600, the PSU is Chieftec CFT-850-14C, I think I didn't mention this but since I have the side cover of the case open I see the PSU label) - but the connected monitor didn't turn off, the power light was on all the time, just changed color to standby mode.

Any ideas what this might be? Might this be some PSU issues? If it will shutdown again I'll try to check it without the UPS.


PS Eh I'm getting tired of testing this PC, it's almost 1.5 month of discovering new issues after previous ones got fixed/stopped occurring