GPU drastically overheating and shutting off to prevent damage automatically

[JazzW]

Basically I was playing a game one day (Age of Empires) and my Moniter suddenly said no signal. Nothing changed until I restarted my computer (I tried various things.) This happened on multiple occasions with multiple games. I looked up the problem and proceeded to test the monitor, the cord between the computer and the monitor, could, power supply, etc. Until I tested the GPU. During the stress test (which lasted about five minutes) I was looking up information about the GPU overheating temperatures and found out that many GPUs turn off if they reach 105 degrees Celsius. In any case, it did heat up to 104 degrees Celsius and seemed about to go up to 105 degrees Celsius and the problem occurred again.

Assuming that this was the problem I looked up ways to keep it cooler. I dusted my PC thoroughly and checked the fans, which were running at full speed. I also tried underclocking the PC to see what it would do. Nothing really worked, although they raised to the time of the stress test from about 5 minutes till crash to about 6 minutes til crash.

I'm now thinking it is a heat sink problem or something along those lines. However it seems it is a bit more expensive to replace that then you would think, so I was planning on just replacing the thermal paste. But I really don't know how much that would really do. As well, I figured someone else might have an idea I could try.

As for my computer, I believe it is custom built by my father about 6 years ago. It is windows 10, the GPU is a GeForce GT 430. Just to be clear I have tested and found no problems with the fans, the CPU, the power supply, the monitor, the connecting cable (can't think of the name at the moment), and various potential causes due to software issues (malware, bugs, etc) . I have updated every last driver on my system. I have gone onto the Nvidia site for my GPU and installed the drivers directly from there, albeit I did not uninstall the old beforehand if that makes any difference.

If you need any more info just ask.

Thanks!

Jazz

 

[Tmas_]

From what you're telling me, it's an old machine with an old GPU. One random day temps spike. What seems to me is that there may be a burst or hole in your heatsink of the GPU. Maybe get a flashlight, take it out the GPU, take off the fan's cover and snoop around. If you don't see any scratches or holes then I have no clue what it is.

 

[pete_101]

Change the paste, if you've blown the dust out and the fan is running, then it has to be the paste that's the problem. Be sure to clean all the old paste off before applying new.

 

[Nighterlev]

Changing the paste likely won't fix those extremely high temps, but from what your telling us it could be because of how old that GPU actually is (GT 430, that's older then any 5-6 year old today)

The only problem though is the temperatures are extremely high, and that GPU will shut the computer down to keep it from cooking it-self. Where do you live? Make sure the computer is open case and not closed, as closed case will only increase the temps even higher.

I would recommend getting rid of the old Nvidia drivers though, but no update is going to fix those type of temps without 1st getting rid of all the dust and such (which you apparently already have)

As said before though, your best bets are to just get a new GPU instead, as that GT 430 just gets way to hot to even be usable from the looks if it. You don't have it over-clocked or anything..do you?

 

[JazzW]

Do, I don't have it overlocked at all. In fact I tried underclocking to see if it would lower the heat. But so far people are saying to open the case, check if there is a problem with the heat sink, And try applying a new paste. OK, I'll try those and see what they do.