GTX 770 Overheating Issues

[DG619]

Specs:

i5-4670K (OC'd @ 4.0Ghz)
MSI GTX 770 2GB
MSI Z87 G45 MOBO
8 GB RAM
Corsair H100i
PC Power & Cooling Silencer Series 750 Watt 80+ Gold PSU

Recently my PC has been running a bit weird. Games that normally run perfect have been stuttering despite showing 60+ FPS. I decided to first try a clean install of Windows 10 since I was still running Windows 7 and it's been a while since I did a clean wipe. At first it seemed better but I think it was just my mind playing tricks on me because today I noticed my temps shooting through the roof. It also seems like the fan sensor isn't picking up the fan RPM properly. I can see the fans spinning a lot faster than the 30-270RPM averages that it is showing in both Afterburner and GPU-Z. I'm worried I may have fried the card. As soon as I boot up a game, even an 8-bit indie title the temps shoot up over 100c. Typically while playing a game the fan RPM shown in GPU-Z and MSI AB will drop down to an alternating 60/90 pattern even though I can see and feel with my hands that they are running faster.

When I'm not playing a game the temps are okay (though I'm not sure if 44c if high for idle)

http://i.imgur.com/9KSKx2I.png

When a game is launched it shoots up over 100c almost immediately forcing me to close the game.

(EDIT: Opened a game long enough to take this screenshot, this was taken almost immediately after launching Pixel Piracy on steam, which is an 8-Bit style indie game)

http://i.imgur.com/bI9Clit.png

The card is clean and free of any dust or blockages in the heat sink/fans though when I first pulled it out there was a good amount on the heatsink. (I can admit I'm not the best at keeping my case clean)

Is a new GPU the only option at this point? Is it possible the fans are just a little tight and new some sort of lube or better cleaning?

(EDIT 2: I'm almost certain the GPU fans are spinning @ 100% speed despite what MSI AB and GPU-Z are telling me.)

 

[Mussels]

what thermal paste is on the GPU? did you replace that when you dusted out the heatsink?

 

[DG619]

what thermal paste is on the GPU? did you replace that when you dusted out the heatsink?

Whatever thermal paste came on the card is still on there. I simply removed the card to clean it and make sure there were no blockages in the heatsink or fans. Although I did build this PC myself I'm not the best with this sort of thing, reapplying thermal paste may be over my head a bit.

 

[Aenra]

This won't help you, as i lack the technical knowledge to even understand what was done, let alone relate it to you.. but..

Have had the same card, was in fact the one i had until i switched to my current rig. Exact same issues, although just for clarity's sake, let me note that mine was an Asus and that the software i used was Asus' GPU tweak? Or whatever it's called anyway..
Because shipping things to China isn't my number one solution, i took it to someone well versed in electronics whom i also happen to trust a lot. He told me that it was real easy to find the issue, due to excessive heat (let's be honest, these fans are shit for this kind of workload, the positioning itself is shit for this kind of workload too, but anyway) some transistor..microprocessor..component..? Whatever, had broken down?.. detached itself?.. And it in turn affected the way the card read and handled the fans. He fixed/replaced that one thing and everything was fine again, for years. With the same fans.
TLDR, all he did was use his multimeter + re-attaching/replacing said component.

As stated, yes, this won't help you, but if i) nothing here does, ii) you don't want to bake it in case it is something detached, you may eventually wish to consider taking it to someone capable of ascertaining the above.

 

[DG619]

This won't help you, as i lack the technical knowledge to even understand what was done, let alone relate it to you.. but..

Have had the same card, was in fact the one i had until i switched to my current rig. Exact same issues, although just for clarity's sake, let me note that mine was an Asus and that the software i used was Asus' GPU tweak? Or whatever it's called anyway..
Because shipping things to China isn't my number one solution, i took it to someone well versed in electronics whom i also happen to trust a lot. He told me that it was real easy to find the issue, due to excessive heat (let's be honest, these fans are shit for this kind of workload, the positioning itself is shit for this kind of workload too, but anyway) some transistor..microprocessor..component..? Whatever, had broken down?.. detached itself?.. And it in turn affected the way the card read and handled the fans. He fixed/replaced that one thing and everything was fine again, for years. With the same fans.
TLDR, all he did was use his multimeter + re-attaching/replacing said component.

As stated, yes, this won't help you, but if i) nothing here does, ii) you don't want to bake it in case it is something detached, you may eventually wish to consider taking it to someone capable of ascertaining the above.

Thanks for the reply, I think at this point if it is something that needs to be fixed I'll just end up putting that money towards upgrading, it's an old card that's kept me from enjoying some of the newer games to their full potential anyways. I just don't have the $$$ right now so getting this hunk of junk to work for a couple more weeks would be awesome. This thing might see the inside of my oven at some point...

 

[Aenra]

Thanks for the reply, I think at this point if it is something that needs to be fixed I'll just end up putting that money towards upgrading, it's an old card that's kept me from enjoying some of the newer games to their full potential anyways. I just don't have the $$$ right now so getting this hunk of junk to work for a couple more weeks would be awesome. This thing might see the inside of my oven at some point...

We are talking about 5 minutes work (for him) and at a cost low enough for him to tell me "you don't owe me anything".
If even that sounds too much for you, try baking it as an ultimate resort. Don't just throw it away.

And semi off-topic, but since you mentioned it? On an FX 9370 + Asus 770 i could run The Witcher 3 @1080p on max settings, only hairworks disabled. Hardly trash worthy i'd think, but obviously, you'll be the judge of that

 

[Mussels]

well thats your answer - replace the thermal paste.

its not exactly complicated, and its cheap. if you can get toothpaste onto a toothbrush without killing yourself, you'll do fine.

 

[DG619]

A couple questions I feel are worth adding... Would temps shoot up over 100c almost as soon as I launch a game if the thermal paste was bad? Is there anything else that would cause this sort of spike? (Fans are running, never had heat issues before so I don't think it has to do with the case/air-flow)

Also, what would cause both GPU-Z and MSI AB to report slow fan RPM's despite me being able to see that they are spinning fast?

well thats your answer - replace the thermal paste.

its not exactly complicated, if you can get toothpaste onto a toothbrush without killing yourself, you'll do fine.

I'll give it a go tomorrow morning.

I'm worried I have been running this card hot for a while now, possibly months. I've never had a BSOD or any artifacts while gaming, is there a good chance I haven't done any permanent damage?

 

[Mussels]

I'll give it a go tomorrow morning.

I'm worried I have been running this card hot for a while now, possibly months. I've never had a BSOD or any artifacts while gaming, is there a good chance I haven't done any permanent damage?

modern cards have a thermal throttle to slow them down before permanent damage occurs. replace the paste and it'll be fine.

 

[Vya Domus]

Thermal paste doesn't suddenly go bad , what can happen is the contact between the chip and cooler can brake due to some physical shock like trying to pull on the cooler. Those unusual fan speeds is a big warning sign to me. It doesn't matter if you can see the fans spinning it may really run at low speeds for some reason.

 

[Mussels]

Thermal paste doesn't suddenly go bad , what can happen is the contact between the chip and cooler can brake due to some physical shock like trying to pull on the cooler. Those unusual fan speeds is a big warning sign to me. It doesn't matter if you can see the fans sniping it may really run at low speeds for some reason

when he cleaned out the heatsink for dust he could have knocked the heatsink and 'broken the seal' on the thermal paste, cracking it and letting air pockets form. old dry paste definitely behaves that way.

 

[DG619]

when he cleaned out the heatsink for dust he could have knocked the heatsink and 'broken the seal' on the thermal paste, cracking it and letting air pockets form. old dry paste definitely behaves that way.

I was having the heat issue before removing the card the cleaning it, that's what actually led me to do that.

Thermal paste doesn't suddenly go bad , what can happen is the contact between the chip and cooler can brake due to some physical shock like trying to pull on the cooler. Those unusual fan speeds is a big warning sign to me. It doesn't matter if you can see the fans sniping it may really run at low speeds for some reason.

I take my PC to work on days I know I'll be there 10+ hours because it gets slow and I can get in some gaming. I stopped about a month ago out of fear that moving it around so much was going to mess something up. I guess it's possible something like that happened but this PC has also seen heavy usage and it gets extremely hot in my room in the summer so I also wouldn't be surprised if the thermal paste has dried up overtime. I built this PC well over 2 years ago.

 

[Vya Domus]

when he cleaned out the heatsink for dust he could have knocked the heatsink and 'broken the seal' on the thermal paste, cracking it and letting air pockets form. old dry paste definitely behaves that way.

Could be.

I was having the heat issue before removing the card the cleaning it, that's what actually led me to do that.

I take my PC to work on days I know I'll be there 10+ hours because it gets slow and I can get in some gaming. I stopped about a month ago out of fear that moving it around so much was going to mess something up. I guess it's possible something like that happened but this PC has also seen heavy usage and it gets extremely hot in my room in the summer so I also wouldn't be surprised if the thermal paste has dried up overtime. I built this PC well over 2 years ago.

So you should replace the TIM then , nothing too difficult just make sure you put enough of it to cover the entire die but obviously not too much and be careful if there are any thermal pads to not tear them apart and leave them exactly where they were. As you take the card apart watch out for the fan header and always use diagonal patterns when you screw down the cooler.

 

[peche]

Thermal paste doesn't suddenly go bad , what can happen is the contact between the chip and cooler can brake due to some physical shock like trying to pull on the cooler. Those unusual fan speeds is a big warning sign to me. It doesn't matter if you can see the fans spinning it may really run at low speeds for some reason.

+1 when op was trying to clean the rig could also broken the bond between cooler and gpu chip, also when pulling the card out,
So replacing TIM will be the best way to find out, its pretty easy, now Youtube its full of videos of a dude repasting your exact video card model, cooler take out in video and all you need...

Specs:

i5-4670K (OC'd @ 4.0Ghz)
MSI GTX 770 2GB
MSI Z87 G45 MOBO
8 GB RAM
Corsair H100i
PC Power & Cooling Silencer Series 750 Watt 80+ Gold PSU

Fill Sys Specs here, fill as much as your knowledge lets you, ask if need help. please try to add exact models, look for models on hardware boxes if you have them arround,

Regards,

 

[Vario]

Agreed Remount HS. I have a 770 runs idle 24*C load 70*C with a big air cooler. Stock was 30*C idle 81*C load. Really easy to dislodge the heatsink. I have done it many times on mine because the holes are too generous relative to the heatsink's bolts. Don't use the machine until after you remount the heatsink. You can use your onboard video in the meantime.

1) Turn off machine
2) detach monitor cables from videocard
3)remove PCI-E power, unscrew and unclip the videocard from the slot.
4) Remove the heatsink by the screws in the back.
5) Take some isopropyl alcohol on a clean rag and wipe the graphics processor unit completely clean of thermal paste.
6)Wipe the thermal paste completely clean off the heatsink's mounting plate with alcohol.
7)Take a pea sized amount of thermal paste and put in center of the graphics processor.
8) Put your finger in a plastic baggy and spread the paste until it a smooth flat sheet of grey on top of the processor. You don't want it thick. Just a thin uniform layer.
9) Carefully reapply the heatsink so that it doesn't smudge the paste but clamps down on top.
10)Screw the heatsink to the same screw tension it had when it was removed.

Reinstall in your machine.

 

[eidairaman1]

Whatever thermal paste came on the card is still on there. I simply removed the card to clean it and make sure there were no blockages in the heatsink or fans. Although I did build this PC myself I'm not the best with this sort of thing, reapplying thermal paste may be over my head a bit.

Take the card apart, replace the thermal compound on the gpu die, thermal pads on memory and vrm circuitry, clean all of your fans and heatsinks of dust. How old is that psu?

 

[Vayra86]

A couple questions I feel are worth adding... Would temps shoot up over 100c almost as soon as I launch a game if the thermal paste was bad? Is there anything else that would cause this sort of spike? (Fans are running, never had heat issues before so I don't think it has to do with the case/air-flow)

Also, what would cause both GPU-Z and MSI AB to report slow fan RPM's despite me being able to see that they are spinning fast?


I'll give it a go tomorrow morning.

I'm worried I have been running this card hot for a while now, possibly months. I've never had a BSOD or any artifacts while gaming, is there a good chance I haven't done any permanent damage?

When you run the card in your system, are all the fans spinning? And what is the fan speed tachometer readup on those fans? Are they spinning at the same RPM?

Check this by adding the fan tachometer speeds to your OSD in Rivatuner, run something that makes your temps go up, and observe. Also have a look at the card itself in your case and check if there's a 'wobble' in the spinning fans.

If you worry about the 100 C temp, you could lower your GPU clock in Afterburner prior to running this test. Nudge it down a good 100 mhz and you should be safe to at least see something going on.

Reason I say this: it generally takes longer for thermal paste to dry out, than it could take to replace a fan. And fans are easily damaged, you could have damaged one during cleaning - not so with the paste.

 

[peche]

 

[INSTG8R]

Before tearing the card apart I would use Afterburner and just set a very aggressive fan curve and see if that helps/changes anything seeing as you have reported odd fan speed readings. Heck put them to 100% and game a little and watch your temps.

 

[eidairaman1]

Never came back

 

[jboydgolfer]

the card exploded. moment of silence please

 

[Killerdroid]

I think it chose self cremation.