|
Problem description: I have been running my PC in its current configuration problem-free for the last ~4 years. In this last month it has been bluescreening with stop code 'VIDEO TDR FAILURE', and shows nvlddmkv.sys as the source of the problem. These BSODs generally occur when I'm playing videogames (War Thunder the main culprit), but I've been able to have long sessions of Rocket League and even Assetto Corsa VR (which renders to both Oculus and my main display) with no problems. I've seen it BSOD only minutes after booting up during general desktop/browser activity, but most days I can spend an entire 8 hour shift working from home with no problems. All the BSODs have the same error and list the same nvlddmkv.sys file as the source Attempted fixes: Google linked me to discussions of other people having the same problem. Following their directions I have - reinstalled my video drivers - reinstalled my video drivers in safe mode - deleted my video drivers using DDU in safe mode, then reinstalled my video drivers - Ran a CHKDSK, no errors - Ran a memory check (mdsched.exe), no errors - Changed the power management mode for my Nvidia card from 'Adaptive' to 'Performance' - Turned down my graphics settings in War Thunder - GPU temp maxes out at around 82c which I've been advised is normal for this card. BSODs have occurred at both ends of the temp range The one recommended fix that I have not been able to complete is to rename the nvlddmkv.sys file to nvlddmkv.sys.old or similar and replace it. No matter what I do my system will not let me rename the file. I get the user access control dialog and say yes, but still get an error saying that I don't have permission. This is a personal computer with no other users besides myself, and I can make Administrator level changes everywhere else. I even tried modifying the Security properties for my non-admin account to have Full Control of this file but I am still denied access to change the files name. Recent changes: None -- Operating system: Win10 Pro 64bit 19044.1706 System specs: - Asus GeForce GTX 1080 Ti Turbo 11GB - MSI S1151 ATX Z370 GAMING PRO CARBON DDR4 Motherboard - Intel Core i7 8700K Six Core LGA 1151 3.7 GHz - x2 Corsair Vengeance LPX 16GB (1x16GB) DDR4 DRAM DIMM - Thermaltake Tough Power 1200W (~2013) Location: Australia I have Googled and read the FAQ: Yes. I thought I had a spare GPU from my old PC lying around that I could test but I think I tossed it. I'm fine with the idea of buying a new GPU but I wish I could think of some way to confirm if it really is a hardware problem first
|
#
?
Jun 4, 2022 05:48
|
|
|
|
| # ? May 1, 2024 11:06 |
|
|
maxe posted:I thought I had a spare GPU from my old PC lying around that I could test but I think I tossed it. I'm fine with the idea of buying a new GPU but I wish I could think of some way to confirm if it really is a hardware problem first Yes, usually when a computer has been running fine for years and BSODs start appearing it means that some piece of hardware is having an issue. You could run http://memtest.org/ overnight at some point to further rule out your RAM as being an issue. You can also run http://hdsentinel.com/ and see if it finds any issues with your drive(s). If those are both okay then you could try using onboard video temporarily and see if things stabilize. If they do then that probably means the GPU or PSU has an issue. If the PSU is from 2013 then it wouldn't be surprising if it's starting to fail (or has trouble powering the system).
|
|
#
?
Jun 4, 2022 23:09
|
|
|
Thrilling conclusion, I borrowed a friend's old Radeon card and hooked it up for testing. After being unable to replicate the BSOD I reconnected my GPU and the error has not recurred since. Maybe there was a shakeup with the drivers as a result of the new card, or maybe the hardware swap process improved a connection somewhere? I have no idea
|
|
#
?
Jun 16, 2022 04:44
|
|