APPCRASH Event 1000 - dwm.exe

For the last couple of weeks i've been having a persisting problem with windows seemingly crashing to a black screen whenever i play 'graphically demanding' games.

After about 4-5 minutes of gameplay (not counting being sat in a menu or paused), the monitor(s) will cease receiving input and turn themselves into the standby state, my graphics card's lighting will turn to default white and the fans will ramp up to 100%. Most of the time i can still hear audio and i can still retain control of any applications through the mouse and keyboard, although the screen(s) will be black and will not receive any input until i power the system off manually via the power button.

After trawling through event viewer i located the suspect events, an event with the code 1000 (or sometimes 1001) relating to an appcrash with dwm.exe, something i understand is a core windows process dealing with desktop windows and their effects, this in turn causes the NVIDIA drivers to crash, leading to a blacked out screen from which it will not recover.

My first fix naturally was to re-install the graphics drivers and did so via the GeForce experience application, to no effect.

Next i rolled back the drivers to a build before the issue arose. Again, no change. The system still crashed to a black screen after gameplay.

From there i decided to disable windows aero effects, the desktop slideshows and whatnot. I also re-installed the offending programs to see if that made any difference. Nothing.

I changed the windows power plan to high-power mode, removed any unused USB devices save the mouse and keyboard, used just the one monitor, reseated every component and cleared the ports for dust and debris, turned down the in-game graphics settings (which should have been my first move but it escaped me until then), ensured windows was up to date, and tried once more .... black screen.

Went through and deleted everything related to Nvidia and it's drivers, uninstalled via DDU and re-installed the older drivers manually via their website. Deleted any program that allowed altering of the GPU RGB settings/fan control and such, including monitoring software and tried again. Interestingly this allowed me to play for a good hour without a crash, and i suspected i could have gone longer but after rebooting the system, it happened again. I ran a Furmark test to see if the GPU would crap out but after twenty minutes on a rather demanding stress test, nothing happened, i only stopped as my temperatures were getting high.

Strange to think it can handle a stress test but not in an actual game.

Nevertheless i pressed on. I tested it with only one stick of RAM in the PC, and went through each systematically, the crash occurred with all of them.

I went and wiped my current windows user profile, re-installed windows afresh and installed nothing but the graphics drivers for my card and the game as a testbed, as i could reliably replicate the crash within a short time frame. It happened once more.

I started to suspect my hardware was an issue so i tested it out with a friends GPU in the system and it didn't crash, though after about 40-50 minutes, the system gave me a BSOD with an IRQL_NOT_LESS_OR_EQUAL error code.

That leads me up to now, i figured it could be the PSU, it's getting on four years old though has not shown any signs of wear or tear recently, though components, like people age and i can understand that. Though i don't have a spare at hand at the moment to test with.

As for the errors, they are almost always followed by a LiveKernelEvent with a code of 141, or sometimes 117, rarely. Looking in windows reliability monitor, there can be as little as one 'Desktop Windows Manager - stopped working' entries, from what i understand windows is trying to reboot the drivers to no avail.

Unfortunately, since resetting windows, most of the logs i have are gone and the only ones are recently. Though i have a couple of entries from one of the first few times it has happened.



https://drive.google.com/file/d/1BKPjCoTNtWst8urincWTinNDJ-Nr3sD_/view - Google Drive to a reliability report XML.


Attached is a DxDiag from before i swapped out the GPU for another one. Though i can post that too if you want. Any other information i'd be happy to provide.

Ok.

PSU is a Corsair CP-9020061 CXM 750w Bronze. Going on 5/6 years old now, not shown any signs of wear and tear though i guess the system has seen rigorous use in it's lifetime.

Speccy - http://speccy.piriform.com/results/0xYD840lqdrfPNvcwPE05Jh

Since resetting windows there were no .DMP files, so i ran through the motions to trigger a crash and uploaded the DMP soon after.

Indeed the outdated Power Supply is an issue in itself and i've already ordered a new one, as it's apparent that it's way past due it needs replacing with a more up to date one. I've already made headway in clearing out lots of unwanted files and cleared much more space on the overstuffed drives, again an issue with me hoarding files I don't need. I've changed the powerplan to Balanced also.

As for the chipset drivers, I've updated everything that was provided via the manufacturers site save for the BIOS, my work schedule is busy and i'm yet to do it.

As for the RAM speeds, i was told that the RAM was fine for the CPU but nevertheless i've set them to run at 2133Mhz via the bios, though interestingly my CPU and MoBo temps have jumped by about 10 degrees.

I did another crash test just for posterity and it now crashes faster than it did before, i could play for 5 minutes or so but now it takes barely 1 minute after loading into gameplay.

I'll upload a more up to date speccy and the crash dumps after i've updated the BIOS sometime in the next day or so.