PC shuts down while playing graphics-intensive games

wndrbr · Post by **wndrbr** » October 31st, 2023, 02:05

PC shuts down while playing graphics-intensive games, event log says "Display driver nvlddmkm stopped responding and has successfully recovered". At first I thought it was overheating, but tried checking temperatures and it was fine. PC specs: ryzen 7 3700x, RTX 3070ti, 32gb ram (two 16gb sticks), win10.

Googling gave me a bunch of possible explanations, but none of them worked so far. Tried deleting (using DDU) and reinstalling video drivers, tried installing old drivers. Checked PSU, reorganized my cable management, removed some dust just in case - it didn't help. Someone said that fast startup could cause issues, I tried disabling it and it didn't help.

Has anyone encountered something like this?

wndrbr · Post by **wndrbr** » October 31st, 2023, 02:08

I've read that enabling user permissions to full control for nvlddmkm.sys may fix this issue... but nvlddmkm.sys is missing in system32.

Post by **rusty_shackleford** » October 31st, 2023, 02:22

run memtest
https://memtest.org/

Rand · Post by **Rand** » October 31st, 2023, 02:30

wndrbr wrote: ↑ October 31st, 2023, 02:05
PC shuts down while playing graphics-intensive games, event log says "Display driver nvlddmkm stopped responding and has successfully recovered". At first I thought it was overheating, but tried checking temperatures and it was fine. PC specs: ryzen 7 3700x, RTX 3070ti, 32gb ram (two 16gb sticks), win10.

Googling gave me a bunch of possible explanations, but none of them worked so far. Tried deleting (using DDU) and reinstalling video drivers, tried installing old drivers. Checked PSU, reorganized my cable management, removed some dust just in case - it didn't help. Someone said that fast startup could cause issues, I tried disabling it and it didn't help.

Has anyone encountered something like this?

Yes, but on an old PC.
It was the PSU that was the problem.

FPYockey · Post by **FPYockey** » October 31st, 2023, 03:05

I had this issue and it was my graphics card dying thats probably what it is.

agentorange · Post by **agentorange** » October 31st, 2023, 03:09

computer is trying to kill itself because you keep playing so much slop on it

aweigh · Post by **aweigh** » October 31st, 2023, 03:19

wndrbr wrote: ↑ October 31st, 2023, 02:05
PC shuts down while playing graphics-intensive games, event log says "Display driver nvlddmkm stopped responding and has successfully recovered". At first I thought it was overheating, but tried checking temperatures and it was fine. PC specs: ryzen 7 3700x, RTX 3070ti, 32gb ram (two 16gb sticks), win10.

Googling gave me a bunch of possible explanations, but none of them worked so far. Tried deleting (using DDU) and reinstalling video drivers, tried installing old drivers. Checked PSU, reorganized my cable management, removed some dust just in case - it didn't help. Someone said that fast startup could cause issues, I tried disabling it and it didn't help.

Has anyone encountered something like this?

Either dying GPU, PSU or motherboard.

Last time I had this start happening it escalated to shutting off outside of games as well, and then eventually shut off permanently. Switched out the PSU with an old one I had laying around but no joy, put in new RAM sticks I bought but no joy (the motherboard gave the error code for RAM, btw); so I took it to a Crispy shop and the guy said it was the motherboard. Replaced the motherboard and put in all of the original components (so NOT the new RAM I bought, the original RAM) and voila, everything still working to this day.

Having an old PSU to quickly test whether or not it was the PSU is a HUGE time-saver. Can literally save you possibly days of troubleshooting simply being able to immediately tell whether it's the PSU or not. See if you can't borrow a PSU from someone else? After eliminating the PSU from the list of possibilities you're left with RAM and motherboard.

And, at least from my experience that I just described, motherboard error codes are useful except if it's the motherboard that's dying, then it spits out wrong codes. Like I mentioned in my case the mobo started giving out the bad RAM code, but it was just garbage output.

wndrbr · Post by **wndrbr** » October 31st, 2023, 03:27

I guess i'm just going to kill myself.

Post by **rusty_shackleford** » October 31st, 2023, 03:33

you can easily test if it's the PSU by using furmark
if it's the PSU your PC will shut down pretty much as soon as it starts

wndrbr · Post by **wndrbr** » October 31st, 2023, 03:35

Oh, forgot to mention. I tried furmark, it worked perfectly fine.

Val the Moofia Boss

In 2021 my AMD RX 570 began to malfunction and I began experiencing GPU crashes for any game of moderate graphical intensity, such as Mass Effect 1 or FF14, or emulating a PS1 or SNES game on RetroArch at 4k with CRT shaders applied and using turbo mode. I then upgraded to a Nvidia RTX 3070 and haven't had any problems since.

Shillitron · Post by **Shillitron** » October 31st, 2023, 03:35

wndrbr wrote: ↑ October 31st, 2023, 02:05
PC shuts down while playing graphics-intensive games, event log says "Display driver nvlddmkm stopped responding and has successfully recovered". At first I thought it was overheating, but tried checking temperatures and it was fine. PC specs: ryzen 7 3700x, RTX 3070ti, 32gb ram (two 16gb sticks), win10.

Googling gave me a bunch of possible explanations, but none of them worked so far. Tried deleting (using DDU) and reinstalling video drivers, tried installing old drivers. Checked PSU, reorganized my cable management, removed some dust just in case - it didn't help. Someone said that fast startup could cause issues, I tried disabling it and it didn't help.

Has anyone encountered something like this?

Are you using SLI in {current_year} ?

GhostCow · Post by **GhostCow** » October 31st, 2023, 03:53

Shillitron wrote: ↑ October 31st, 2023, 03:35

wndrbr wrote: ↑ October 31st, 2023, 02:05
PC shuts down while playing graphics-intensive games, event log says "Display driver nvlddmkm stopped responding and has successfully recovered". At first I thought it was overheating, but tried checking temperatures and it was fine. PC specs: ryzen 7 3700x, RTX 3070ti, 32gb ram (two 16gb sticks), win10.

Googling gave me a bunch of possible explanations, but none of them worked so far. Tried deleting (using DDU) and reinstalling video drivers, tried installing old drivers. Checked PSU, reorganized my cable management, removed some dust just in case - it didn't help. Someone said that fast startup could cause issues, I tried disabling it and it didn't help.

Has anyone encountered something like this?
Are you using SLI in {current_year} ?

You can't even do SLI with a 3070. SLI is completely dead now.

Rand · Post by **Rand** » October 31st, 2023, 07:52

aweigh wrote: ↑ October 31st, 2023, 03:19

wndrbr wrote: ↑ October 31st, 2023, 02:05
PC shuts down while playing graphics-intensive games, event log says "Display driver nvlddmkm stopped responding and has successfully recovered". At first I thought it was overheating, but tried checking temperatures and it was fine. PC specs: ryzen 7 3700x, RTX 3070ti, 32gb ram (two 16gb sticks), win10.

Googling gave me a bunch of possible explanations, but none of them worked so far. Tried deleting (using DDU) and reinstalling video drivers, tried installing old drivers. Checked PSU, reorganized my cable management, removed some dust just in case - it didn't help. Someone said that fast startup could cause issues, I tried disabling it and it didn't help.

Has anyone encountered something like this?
Either dying GPU, PSU or motherboard.

Last time I had this start happening it escalated to shutting off outside of games as well, and then eventually shut off permanently. Switched out the PSU with an old one I had laying around but no joy, put in new RAM sticks I bought but no joy (the motherboard gave the error code for RAM, btw); so I took it to a Crispy shop and the guy said it was the motherboard. Replaced the motherboard and put in all of the original components (so NOT the new RAM I bought, the original RAM) and voila, everything still working to this day.

Having an old PSU to quickly test whether or not it was the PSU is a HUGE time-saver. Can literally save you possibly days of troubleshooting simply being able to immediately tell whether it's the PSU or not. See if you can't borrow a PSU from someone else? After eliminating the PSU from the list of possibilities you're left with RAM and motherboard.

And, at least from my experience that I just described, motherboard error codes are useful except if it's the motherboard that's dying, then it spits out wrong codes. Like I mentioned in my case the mobo started giving out the bad RAM code, but it was just garbage output.

I had a motherboard die without codes.
It turns out that several of the big caps that seem to be regulating power to the CPU went bad.
When I looked, they were a bit bulgey on the tops and bottoms.

gerey · Post by **gerey** » October 31st, 2023, 07:53

Might be silly, but have you tried checking if the card is slotted properly? Or simply try taking it out of the slot and put it back in? I've had a friend whose card kept crashing his system under heavy loads at random and after extensive troubleshooting and finding nothing wrong, he tried taking the card out and putting it back in, ensuring it was properly set, and it "magically" fixed his issues.

wndrbr · Post by **wndrbr** » October 31st, 2023, 08:01

gerey wrote: ↑ October 31st, 2023, 07:53
Might be silly, but have you tried checking if the card is slotted properly? Or simply try taking it out of the slot and put it back in? I've had a friend whose card kept crashing his system under heavy loads at random and after extensive troubleshooting and finding nothing wrong, he tried taking the card out and putting it back in, ensuring it was properly set, and it "magically" fixed his issues.

yeah i tried that. It slotted properly, pci-e lock clicked. The card itself is really damn big and heavy, maybe i should buy an anti sag bracket.

Rand · Post by **Rand** » October 31st, 2023, 08:04

wndrbr wrote: ↑ October 31st, 2023, 08:01
maybe i should buy an anti sag bracket

Deffo get a bracket or adjustable post.

wndrbr · Post by **wndrbr** » October 31st, 2023, 11:24

rusty_shackleford wrote: ↑ October 31st, 2023, 02:22
run memtest
https://memtest.org/

ran two passes, no errors.

Pipeweed · Post by **Pipeweed** » October 31st, 2023, 11:51

is there a coredump? it explicitly says its a driver issue? i'd test on linux

Emphyrio · Post by **Emphyrio** » October 31st, 2023, 13:05

I didn't get this error message but I had similar symptoms and the solution was to run the windows update repair tool.

maidenhaver · Post by **maidenhaver** » October 31st, 2023, 14:18

The only solution is renounce gaming forever. Look at it as a new beginning, you had shit taste, anyway.

Sweeper · Post by **Sweeper** » October 31st, 2023, 14:59

It's 100% the GPU, it's always the fucking GPU.

somerandomdude · Post by **somerandomdude** » October 31st, 2023, 17:14

Did you check memory junction temps and hotspot temps on your GPU with HWiNFO64?

On older GPUs the memory junction or hot spot temps can be considerably hotter than core temps.

somerandomdude · Post by **somerandomdude** » October 31st, 2023, 17:34

You can try under volting your card with MSI curve editor. I've done this on several GPUs, and it works pretty well.

You'll lower temps and power draw, if what you're experiencing is related to either, this might fix it.

Slavic Sorcerer · Post by **Slavic Sorcerer** » October 31st, 2023, 17:52

I've had something like this for a year

Turns out my graphic card was at fault. Not because of specification but something went wrong with it. It worked, but was struggling with games having intensive graphics
Sometimes it would work on a blessed day, but more than often it just shutting down the PC and booting it again

It might be difficult, but if you have someone that would lend your their graphic curd for a test run, you can see if there is something wrong with the graphic card itself

Before that I wiped out the system and installed everything anew, thinking it's software issue (cleaning didn't help and other physical examinations proved nothing) - all in vain

GhostCow · Post by **GhostCow** » October 31st, 2023, 19:34

It could still be the PSU, even if it doesn't fuck up in furmark, but I doubt it. Ampere is know for big random power spikes that a lot of power supplies can't handle. Personally I would try wiping the ssd and reinstall everything from scratch. That's always my last resort and it usually works.

BENEFACTOR · Post by **BENEFACTOR** » November 1st, 2023, 00:22

Maybe overheating? Never had problems like that with any MAC i've owned, only windows computers.

wndrbr · Post by **wndrbr** » November 5th, 2023, 08:40

Reduced the amount of crashes by switching to "maximum performance" in nvidia control panel. During the last four days it only crashed once.

GhostCow · Post by **GhostCow** » November 5th, 2023, 13:46

wndrbr wrote: ↑ November 5th, 2023, 08:40
Reduced the amount of crashes by switching to "maximum performance" in nvidia control panel. During the last four days it only crashed once.

This is a common common fix and all of us in this thread are retarded for not thinking of it.

Lutte · Post by **Lutte** » November 7th, 2023, 23:40

GhostCow wrote: ↑ November 5th, 2023, 13:46

wndrbr wrote: ↑ November 5th, 2023, 08:40
Reduced the amount of crashes by switching to "maximum performance" in nvidia control panel. During the last four days it only crashed once.
This is a common common fix and all of us in this thread are retarded for not thinking of it.

Eh, to me it sounds like the GPU might have some damage and changing this setting just masks it for a while before it gets aggravated. If it ran fine for months without any issue without using the max performance profile, and it suddenly acted up recently, and doing this fixes it, it's possible that what triggered the issue in the electronics is fluctuation of temperatures causing mechanical stress (thermal expansion and contraction cycles) and that using the max perf profile causes the GPU to settle around a certain high temperature, making it go through fewer of said cycles during gameplay - whereas a normal or economy power profile will fluctuate more heavily depending on the scenes as the GPU powers down whenever it can afford to. Of course, for thermal stress - within acceptable ranges, to cause issues, it means the hardware has a fault to begin with.

It's /never/ normal to have to mess with power profiles just to stop crashes from happening. A healthy computer doesn't constantly crash out of the fucking box.