Question System Freeze

atticus_maytrap

Reputable
Nov 12, 2018
74
3
4,545
Hi there,

so this is a problem i thought had been fixed but has popped back up again.

In late February - right after an Nvidia driver release - i started to get system freezes, as in mouse not moving, nothing responding, hard restart needed. Screen did not go to black, just remained frozen in place, and any sound coming from the machine (music, video sound etc.) kept on playing perfectly, however, these freezes only ever happened in a very specific situation. Namely, i use 3 monitors for work (architecture) but then switch that back to one monitor for gaming. The system freeze only ever occurred right after i turned the second and third monitors off.

The freezes then got progressively worse / seemingly more random (never under load though, usually right after gaming), until the machine wouldn't even boot to windows. I unplugged the main monitor from the graphics card (MSI 1080Ti) and into the mobo, system back up, and then on advice from the Nvidia forum, rolled back to an older driver using the DDU in safe mode method.

Everything has been running smoothly since then (3rd of April) and then last night, right as i was finishing up work and turning off the extra monitors BAM - another system freeze, right after the 2nd and 3rd monitors get turned off.

Since the first round of problems i had installed whocrashed and got back the following message:

The following dump files were found but could not be read. These files may be corrupted:
C:\WINDOWS\LiveKernelReports\PoW32kWatchdog-20210421-2243.dmp

ive run chkdsk, sfc /scannow and the windows memory diagnostic tool, all came back clean.
Is there anything else i can do here short of the nuclear option?

note: one theory i have is the power settings for the graphics card. I have this set to adaptive, and the freezes all seem to occur when the card is "powering down" or moving from a state of higher workload (3 monitors, gaming) to lower workload (1 screen, in the 20-30 second window after gaming), but has yet never actually happened under load (gaming or rendering).
Im going to try switching this over from adaptive to performance so that it maintains power and see if it crashes again. Is this a fools wish?
 
Apr 21, 2021
4
0
10
Hi there,

so this is a problem i thought had been fixed but has popped back up again.

In late February - right after an Nvidia driver release - i started to get system freezes, as in mouse not moving, nothing responding, hard restart needed. Screen did not go to black, just remained frozen in place, and any sound coming from the machine (music, video sound etc.) kept on playing perfectly, however, these freezes only ever happened in a very specific situation. Namely, i use 3 monitors for work (architecture) but then switch that back to one monitor for gaming. The system freeze only ever occurred right after i turned the second and third monitors off.

The freezes then got progressively worse / seemingly more random (never under load though, usually right after gaming), until the machine wouldn't even boot to windows. I unplugged the main monitor from the graphics card (MSI 1080Ti) and into the mobo, system back up, and then on advice from the Nvidia forum, rolled back to an older driver using the DDU in safe mode method.

Everything has been running smoothly since then (3rd of April) and then last night, right as i was finishing up work and turning off the extra monitors BAM - another system freeze, right after the 2nd and 3rd monitors get turned off.

Since the first round of problems i had installed whocrashed and got back the following message:

The following dump files were found but could not be read. These files may be corrupted:
C:\WINDOWS\LiveKernelReports\PoW32kWatchdog-20210421-2243.dmp

ive run chkdsk, sfc /scannow and the windows memory diagnostic tool, all came back clean.
Is there anything else i can do here short of the nuclear option?

note: one theory i have is the power settings for the graphics card. I have this set to adaptive, and the freezes all seem to occur when the card is "powering down" or moving from a state of higher workload (3 monitors, gaming) to lower workload (1 screen, in the 20-30 second window after gaming), but has yet never actually happened under load (gaming or rendering).
Im going to try switching this over from adaptive to performance so that it maintains power and see if it crashes again. Is this a fools wish?

hi
If you want to get back to before the update you could use System Restore (if you’ve created restore points).
This feature has saved me when I’ve had issues after an update.
it’s not a perfect solution but could help stabilize your system until you can narrow down your issue

good luck!
 

atticus_maytrap

Reputable
Nov 12, 2018
74
3
4,545
hi
If you want to get back to before the update you could use System Restore (if you’ve created restore points).
This feature has saved me when I’ve had issues after an update.
it’s not a perfect solution but could help stabilize your system until you can narrow down your issue

good luck!

Hey Swifty, thanks for replying

i've not created a restore point, but i dont think this is related to any specific update, windows or otherwise.
It began with consistency after the late February Nvidia driver update, and rolling back the driver there seemed to clear the problem up, but it has happened in the past - maybe 5/6 months ago i think - and again in that exact same situation, switching from 3 screens to 1.
The difference in refresh rates between the 3 monitors is something ive also considered, my main monitor being 144hz and the other 2 just 60hz