Display driver nvlddmkm stopped responding and has successfully recovered.

[willy bammy]

been getting this error in event viewer lately. Randomly my screen will just go black for a second and come back, usually messing up any video currently on the screen, and this warning pops up in event viewer.
"Display driver nvlddmkm stopped responding and has successfully recovered."
Any clue what this is or how to fix it?

[willy bammy]

Publicado originalmente por Bad 💀 Motha:

Yes if doing any OC what so ever, revert it to stock settings and then do more tests to see if issue still prsent.

Usually if it is faulty ram or psu you'd be getting random reboots or bsod, not a TDR type issue.

Disable all power saving in bios and the OS as this helps also. Plus on a gaming PC all that is best disabled anyways.

The only OC is the factory overclock, I'm hesitant to play with that at all since I've heard messing with the clock settings voids your warranty

[Bad 💀 Motha]

No clock changes can mess with a warranty. Complete rubbish non-sense.

Plus you want to actually turn the OC down, not up, so definitely no harm there.

While rare I did have this very same issue with a factory OCed gtx 580 back even on Win7 64bit. Increasing the TDR Delay fixed it for most of that issue, but some games simply were not fully stable unless I actually down-clocked the gpu back down to near the nvidia reference clocks, which then fixed the issue for said game crashing amd event viewer saying it was driver or TDR timeout related. Games like GTA4 for example

[willy bammy]

Publicado originalmente por Bad 💀 Motha:

No clock changes can mess with a warranty. Complete rubbish non-sense.

Plus you want to actually turn the OC down, not up, so definitely no harm there.

While rare I did have this very same issue with a factory OCed gtx 580 back even on Win7 64bit. Increasing the TDR Delay fixed it for most of that issue, but some games simply were not fully stable unless I actually down-clocked the gpu back down to near the nvidia reference clocks, which then fixed the issue for said game crashing amd event viewer saying it was driver or TDR timeout related. Games like GTA4 for example

oh well the issue doesn't exclusively happen while in game, I'll get randomly just perusing the internet and ♥♥♥♥. Sometimes it'll make my screen go black for a sec, sometimes nothing will happen. If it happens during a game, sometimes it'll crash

[Bad 💀 Motha]

Then that means you have things like a Web browser using your GPU.

In apps like Edge, Chrome, Steam Client... Disable Hardware Acceleration to stop those apps from using gpu. Any i7 is good enough to handle this without the need to have any help from your gpu.

[[☥] - CJ -]

Do you have the NVIDIA Control panel power setting set to Max Performance by any chance?

[Illusion of Progress]

Yeah, sounds like the same exact issue I was having in the late Fermi days and drivers of those times; it was infamous back then too. Playing games could be fine. Browsing the internet could trip a driver crash. Again, pretty sure this was "fixed" for most people simply by using older drivers, and then later newer ones once the issue was ironed out. Sounds like you've been having the issue from the start and either it's getting worse, or you're noticing it more.

With recent news of some RTX 3080 and/or 3090 being made underspec, I really don't have high faith in nVidia hardware or even drivers like I used to. Drivers after 373.06 were messing with Minecraft and other games for me, Resident Evil 2 was having issues with many drivers on and off (a certain cut scene in particular, but also some people having issues with extreme darkness, FWIW I avoided all of these issues on 373.06), etc. Everyone gives AMD flak for this (and to be fair, rightfully so at times) but I've been watching nVidia also get bad here too for the last half a decade or so. Actually, further back than that; GeForce 8800/9x00 through GTX 2x0 series were known to fail due to cracking solder (remember the "bake it" fixes?).

[Pickleman]

.

[SeriousCCIE]

I had this same issue earlier this month. The system crashed and ended up rebooting to a black screen with a mouse cursor.

That was not so easy to resolve. That was caused by, I expect, overclocking the bclk on the motherboard in addition to the card/cpu etc. I was attempting to regain lost glory; after adding some peripherals I had to reduce the bclk speeds to get the system stable, and I was trying to get back to close to that.

In my system at least, the bclk speed is less of a boost for the CPU as it is for the underlying storage system and RAM/memory performance. If I have it set too high, I'll get NVMe storage presentation errors and drives will just disappear from the OS and the system will hang. If the game in question is using shader caches on the same drive that received a PCIe bus reset like that... I get the same error the OP had.

In my situation, the PSU wasn't a contributing factor, or at least not directly. It may be that the PCIe slots are not receiving enough power appropriately due to the amount of stuff connected, but the PSU itself is not struggling--it's more of a demanding too much of the system problem related to aggressing OCing.

(Incidentally, I used an EVGA pcie slot 12v power booster and was able to eliminate the pcie bus reset problem for that half of the motherboard slots -- I am going to add a second one later on and see if I can then increase my bclk back to what it was before -- but heck you get to a point you just want it to work and are done with tweaking!)

[willy bammy]

Update on this, today I've been getting another issue in tandem with this one (event 4101).
Event 4109, "Application X has been blocked from accessing Graphics hardware."
Got one saying discord was blocked (discord's entire screen went blank for a few seconds and came back) and for DOOM Eternal (crashed the game).
Pops up with a windows notification as well saying the same thing
Is it likely my graphics card is just corrupted/faulty?

[Illusion of Progress]

Have you tried older drivers?

[DeadPhoenix]

Post PC specs. Remove any OC.

[Bad 💀 Motha]

And launch trusted apps all via Run As Admin. It's a must for game clients. And when a game client is launched via Run as admin, so will any game launched from it without the need to apply a run as admin property to each game exe

Disable the Windows UAC feature, or at least turn the setting down so it never triggers the graphical overlay when UAC pops up

[hawkeye]

It's usually caused by overclocking, even factory overclocking. The gpu is unstable. Use the software that came with card to set it to normal levels and see what happens.

[Bad 💀 Motha]

Just use MSI Afterburner for any and all gpus. You can't go wrong with it and it works on pretty much any amd or nvidia gpu, even mobile ones.

But yes look up the nvidia reference clocks for your gpu model and try those clocks in MSI Afterburner. Just for Core and Memory, don't mess with Power Boost or Voltage.

[willy bammy]

Reviving this, because I still haven't fixed it and it's managing to get worse.
Still constantly getting events 4101 and 4109, even outside of games it'll cause my screen to just go black for a few seconds before coming back. Now I have had a BSOD while playing Street Fighter V. From what I can tell, my computer isn't overheating at all, so I don't know what made this happen, but I figure it is related to this.
I'm not sure what else to do. All I have left to try at this point is a BIOS update, underclock my card (which I'm weary about), or get a replacement GPU/RAM/PSU/whatever is causing it.