[SOLVED] Computer Crashes - GPU Fan Goes Incredibly Loud and Display Turns Off

[Danny2477]

Hi guys,
I've been plagued with this problem for a while. My computer runs fine when it's idle and when I'm browsing the internet, watching videos etc. However, when I begin gaming, my computer randomly crashes/shuts down. There's no time frame on the issue - sometimes it crashed after 20 minutes, other times I can play for 6 hours and have no issues.

As the title says, my screen turns off (no connection available message) and the fan on the GPU begins to spin aggressively and really loud. I've monitored the temperatures of my GPU when playing games and it sits around 80'C - which isn't overheating but isn't the best. I've actually ordered thermal paste to reapply it as a temporary fix to reduce the temperature.

I originally thought it was the GPU overheating, but as I mentioned above the temperature doesn't seem that bad. Indeed, I've ran the FurMark benchmark test and it exceeded 100'C, at which point I turned it off just to make sure I didn't overheat it on 'purpose'.

CPU: Intel Core i5 3570
Motherboard: ASROCK Z77 Pro4
Graphics Card: nVidia GTX 760
RAM: cannot remember specific make but 12GB
PSU: 700W (I think)
2 case fans (both working), GPU fan works and CPU fan works.


I am not sure if this is the error that is causing my GPU to shut down. I found it in EventViewer and attempted to research it but most of the other individuals having the same error say their game crashes, not their computer (?):

The description for Event ID 13 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.

The following information was included with the event:

\Device\Video3
Graphics Exception on (GPC 0, PPC 0): ESR 0x503038=0xffffffff

The message resource is present but the message was not found in the message table


I am at a complete loss as to what to do. I cannot afford a new graphics card as I was saving for an entire new rig, and don't want to spend £160 as a temporary fix to buy a more recent model afterwards.

 

[nishilamin1213]

Hi, Your PSU seems to be fine for the system, and the only thing i can think of is a GPU driver update. Try "DriverEasy" the free version is slow however it should do the job if drivers are an issue.

 

[Danny2477]

Hi, Your PSU seems to be fine for the system, and the only thing i can think of is a GPU driver update. Try "DriverEasy" the free version is slow however it should do the job if drivers are an issue.

Thanks for the quick reply. I'll run it now and test it out. However, I did actually update my nVidia drivers on Tuesday to the latest drivers available (non Beta) and performed a clean install.

 

[Danny2477]

Other errors I have noticed occuring regularly:

The Desktop Window Manager process has exited. (Process exit code: 0x000000ff, Restart count: 4, Primary display device ID: NVIDIA GeForce GTX 760)


Faulting application name: dwm.exe, version: 10.0.18362.387, time stamp: 0x8e064b77
Faulting module name: unknown, version: 0.0.0.0, time stamp: 0x00000000
Exception code: 0xe0464645
Fault offset: 0x0000000000000000
Faulting process ID: 0x1aac
Faulting application start time: 0x01d6090cef7df1a6
Faulting application path: C:\Windows\System32\dwm.exe
Faulting module path: unknown
Report ID: e2fda86f-108b-4299-b35d-b971a6075858
Faulting package full name:
Faulting package-relative application ID:

 

[nishilamin1213]

Thanks for the quick reply. I'll run it now and test it out. However, I did actually update my nVidia drivers on Tuesday to the latest drivers available (non Beta) and performed a clean install.

Seeing as you've updated drivers, that may not be an issue then, I did have a similar issue to you, however that was when i used dual PSU, so mine was expected. Is the GPU hardware in good condition.

 

[MadsModsat]

... GPU driver update. Try "DriverEasy"

I would highly recommend not using that program. It causes more issues than solutions, and it installs a lot of different things to your PC, that you would be much better off without.

Always get the drivers from the source, wheter it is AMD, NVidia, ASUS, MSI and what not - just don't use DriverEasy

 

[nishilamin1213]

I would highly recommend not using that program. It causes more issues than solutions, and it installs a lot of different things to your PC, that you would be much better off without.

Always get the drivers from the source, wheter it is AMD, NVidia, ASUS, MSI and what not - just don't use DriverEasy

Oh, i've used it for a long time without issues, and its only installed itself, it is a bit clunky though, and yes, its not the best program but it does the job. What program would you recommend?

 

[Danny2477]

Seeing as you've updated drivers, that may not be an issue then, I did have a similar issue to you, however that was when i used dual PSU, so mine was expected. Is the GPU hardware in good condition.

In all honesty it's coming up to 7 years old. At a glance/probing into the case the hardware appears to be in good condition. Hasn't received any knocks and was only removed to be transferred into the current motherboard (done about 5 years ago but this issue is fairly recent - potentially spanning back two years but only the last year have I used my desktop again).

I will look at it properly tomorrow when I reapply the thermal paste.

 

[MadsModsat]

Oh, i've used it for a long time without issues, and its only installed itself, it is a bit clunky though, and yes, its not the best program but it does the job. What program would you recommend?

I would never recommend any of the free (or paid) programs that promises you to find outdated drivers and update them for you.

Some programs want to update drivers that are not actually outdated, sometimes you are offered the completely wrong drivers.

You may not be aware of everything DriverEasy installs when you set it up first time, but that's the idea. It installs unwanted software, under the pretence of simplifying something some people find a bit tedious.

I can only recommend going to the the manufacturers website for drivers. Motherboard drivers from the manufacturer, graphics drivers from AMD or Nvidia.

There are plenty of warnings against this type of software.

I have made my advice available, so OP can take it or ignore it. I can only hope OP will google DriverEasy, and skip past the sponsored content, and see what the recommendations are, before deciding on using the software or not

Personally I think it is very bad advice to recommend DriverEasy or similar programs

 

[nishilamin1213]

I would never recommend any of the free (or paid) programs that promises you to find outdated drivers and update them for you.

Some programs want to update drivers that are not actually outdated, sometimes you are offered the completely wrong drivers.

You may not be aware of everything DriverEasy installs when you set it up first time, but that's the idea. It installs unwanted software, under the pretence of simplifying something some people find a bit tedious.

I can only recommend going to the the manufacturers website for drivers. Motherboard drivers from the manufacturer, graphics drivers from AMD or Nvidia.

There are plenty of warnings against this type of software.

I have made my advice available, so OP can take it or ignore it. I can only hope OP will google DriverEasy, and skip past the sponsored content, and see what the recommendations are, before deciding on using the software or not

Personally I think it is very bad advice to recommend DriverEasy or similar programs

Ok, I must have been under the wrong impression, thanks for the advice

 

[MadsModsat]

Ok, I must have been under the wrong impression, thanks for the advice

Just to clarify - I know it was with the best intentions possible, I do not mean to make it sound like you had bad intentions or anything. It is the software I find problematic, nothing else

Also, I could be completely wrong and not know anything, so always consider advise carefully, so you are sure it is not just hot air.

But it is still my advice to stear clear of this kind of programs, or at least reaserch very well before using it, what the program actually says it does, and what it might not tell you (in a clear and direct way, at least)

The sponsored content on Google search praises the software, but looking a bit further, it isn't difficult to find advice against using the program

 

[Danny2477]

So my graphics card appears to look fine. It works perfectly well up until the crash so I don't think there is anything physically wrong with it.

As for my RAM, I viewed it on Task Manager and it appears normal. Not sure how it copes in-game but there's nothing to make me think it's playing up.

Conducting a fresh reinstall of the latest nVidia drivers now to see if the problem rectifies itself but I highly doubt it considering they were freshly installed just two days ago and it happens on previous drivers too.

 

[henterpriser]

The maximum safe GPU temp For your gpu is 93c but mainly the maximum safe temp for all components(CPU GPU etc.) is 85c and you mentioned that the gpu reached 100c(water boiling temp!) using Furmark And this is highly dangerous for your GPU and staying in that temp will damage your GPU over time and may broke it. and you should never let it reach beyond 85c reapplying thermal paste and setting the right curve may fix it.

Rather than those crashes happening, you should optimize your fan curve more aggressively
because GTX 760 is one of those HOT GPUs and i recommend you to put the curve like this.

If your GPU manufacture provided you the software do it using that software.
otherwise use third party apps such as SpeedFan etc.

80c--->100% Fan
70c--->80%
65c--->70%
60c--->60%
55c--->50%
50c--->Lowest Possible If your gpu have 0db technology set it fan to stop otherwise set the minimum fan speed your gpu supports.

Good Luck.

 

[Danny2477]

The maximum safe GPU temp For your gpu is 93c but mainly the maximum safe temp for all components(CPU GPU etc.) is 85c and you mentioned that the gpu reached 100c(water boiling temp!) using Furmark And this is highly dangerous for your GPU and staying in that temp will damage your GPU over time and may broke it. and you should never let it reach beyond 85c reapplying thermal paste and setting the right curve may fix it.

Rather than those crashes happening, you should optimize your fan curve more aggressively
because GTX 760 is one of those HOT GPUs and i recommend you to put the curve like this.

If your GPU manufacture provided you the software do it using that software.
otherwise use third party apps such as SpeedFan etc.

80c--->100% Fan
70c--->80%
65c--->70%
60c--->60%
55c--->50%
50c--->Lowest Possible If your gpu have 0db technology set it fan to stop otherwise set the minimum fan speed your gpu supports.

Good Luck.

The longest I've let it run past 95° on the tests it about 5 minutes but the results stay the same.

Why would you turn off the fan (essentially) when the temperature is below 50°c?

I've attempted to have the fan on 100% when temperatures hit 70 and 80 but to no avail.

 

[henterpriser]

The longest I've let it run past 95° on the tests it about 5 minutes but the results stay the same.

Why would you turn off the fan (essentially) when the temperature is below 50°c?

I've attempted to have the fan on 100% when temperatures hit 70 and 80 but to no avail.

If you already had a 0% technology GPU they drop the fan to 0% while the temperature goes below 60c(55c for rog strix)

And the reason they are doing this is:
Firstly because the temp 50-60 is mainly for idle status and your main target in idle status is using the lowest power/noise during this state as long as your cpu/gpu is not being used.
Second of all the 50-60c temp is like a cold winter for your GPU and the components.

So what's the point of the fans running when there is no use on gpu and its cooling by its's own?

 

[henterpriser]

The longest I've let it run past 95° on the tests it about 5 minutes but the results stay the same.

Why would you turn off the fan (essentially) when the temperature is below 50°c?

I've attempted to have the fan on 100% when temperatures hit 70 and 80 but to no avail.

95c Is OVER YOUR GPU MAX SAFE TEMP provided by Nvidia

And you should do something about it such as throttling down the voltage/core clock Or make the cooling better by reapplying paste or running fans faster or maintain the air Inhale/Exhaust through your Pc Chasis.

Hope your problem fixes.
Good Luck.

 

[Danny2477]

95c Is OVER YOUR GPU MAX SAFE TEMP provided by Nvidia

And you should do something about it such as throttling down the voltage/core clock Or make the cooling better by reapplying paste or running fans faster or maintain the air Inhale/Exhaust through your Pc Chasis.

Hope your problem fixes.
Good Luck.

I'll set my fans to operate much lower at the lower temperatures. When my GPU hits above 85c it begins to throttle automatically. I will be reapplying thermal paste tomorrow, although I do not think it is my GPU overheating that is the issue.

I ran memory diagnostics and they came back clean so no worry there.

I ran a test using Payday: The Heist. I figured an old game wouldn't make my GPU overheat. As confirmed, it never reached above 75c but my desktop still crashed. The error log doesn't show as much information as I've listed above this time either.

 

[henterpriser]

Doing some research found that plenty of those cards starts to do like this as you can see in the link below:
https://forums.evga.com/m/tm.aspx?m=2530081&p=160
they all having the same screen goes black and the fan going to 100% problem and most of them fixed it by returning the card and getting a new one(RMA)!
seemingly it's a hardware problem.
although some managed to fix it without returning the card that you can read there.

Good Luck.

 

[Danny2477]

Doing some research found that plenty of those cards starts to do like this as you can see in the link below:
https://forums.evga.com/m/tm.aspx?m=2530081&p=160
they all having the same screen goes black and the fan going to 100% problem and most of them fixed it by returning the card and getting a new one(RMA)!
seemingly it's a hardware problem.
although some managed to fix it without returning the card that you can read there.

Good Luck.

Damn, sad to read it's a hardware issue that'll most likely need replacing. However, that thread is tailored specifically to EVGA 1080FTW. My graphics is an MSI GTX760. Same problem though, it's just surprising to see so many with the same issue but they are all using the same card.

Don't suppose you have any graphics cards in mind that are reasonably cheap to get me through in the meantime? I've noticed the 1050ti on Amazon is reasonably priced.

 

[henterpriser]

No. Never go for 1050 ti in 2020
With the same price you can go for:

AMD side: RX 570
A bit more expensive: RX 580

Nvidia side: GTX 1650(Super,TI)
A bit more expensive: GTX 1660(Super,TI)

Also Graphic cards are not Cars(unless they mined with it) which means you can get a sightly better price while buying a unmined Used GPU and save some money.
For sure you can do some stability test on it and ensure the card is doing fine otherwise return the card.