GTX980 Overheating Help!

[nrc0]

Hello and thanks for clicking my post!

 

Case - Bitfenix Prodigy

 

Motherboard - Asus Maximus VI Impact

 

CPU - Intel Haswell i5 4670k

 

GPU - EVGA GTX 980 SC

 

Memory - G.Skill Ripjaws X Series 8Gb (2x4gb)

 

Storage - x2 Samsung 850 SSDs (250gb)

 

CPU Cooler - Corsair H60 Liquid Cooler

 

PSU - Corsair HX650 80 Plus Gold

 

OS - Windows 8.1 OEM

 

Some Background...

 

So about a month ago I bought a new EVGA GTX980. It replaced my old Gigabyte GTX760. Note that I had the 760 in my PC for over a year and never experienced any problems.

 

Everything was going great for about a week, then I started running into problems... More specifically my computer crashes while playing games, and only games. I never have a problem when I'm not in a game. Sometimes it crashes to desktop, sometimes it locks my computer with a loud, repetitive audio lock which requires a full computer restart. Below is a list of games that I have experienced multiple crashes on, usually within 15 mins of playing, which render the games unplayable:

 

- Assassin's creed: Unity

- Assassin's creed: Syndicate

- Rocket League

- Rainbow six: Siege (Beta and Full Game)

- Sid Meyer's Civilization 5

- Fallout 4

- Star Wars: The Old Republic

- Counter Strike: Global Offensive

 

The only game I have not experienced this in is League of Legends, which is not a very demanding game in general.

 

Things I have done:

 

- Everything in my system is running default (No overclocks on CPU or GPU)

 

- I set up my second monitor for testing purposes (I only ever have 1 monitor connected for general use and while gaming) and I started up multiple games.

          I had benchmark programs running on the second screen, including EGVA's precision X, RealTemp, GPU-Z, and CPUID Hw Monitor Pro, as well as                       windows 10 task manager to watch memory and cpu usage. All the numbers were running within comfortable specs, the GPU never exceeded temp 80c

          and my cpu was staying around a cool 60-70c. I did multiple trials and everything was normal, but to no avail, I always got a crash 10-30 mins into playing.

 

- I uninstalled Nvidia drivers and installed current ones 2 or 3 times

 

- I updated my motherboard bios, as well as loading defaults on my entire bios

 

- I even reinstalled windows 10 (Saw this on a forum but still didn't work)

 

- I put back in my old Graphics card (Gtx760) and played a few games. Didn't experience one crash

 

- The only thing I changed when I started experiencing these crashes was my graphics card, I didn't change out any other components other than my monitor (but that wouldn't matter right?...RIGHT??)

 

- All my fans in my case are spinning

 

- I have tried both 1080p and 1440p resolution in games with same results

 

- I have not experienced a crash outside of playing a game

 

- None of my games are modded

 

- I took my card out of the case, de-dustified the whole computer, and reseated the card

 

Important Info:

 

Because I have a mini-ITX case, my Graphics card sits right up against my side panel. I decided to do a test.  

 

I took off my side panel, exposing the card and its 2 fans to fresh air, and I played some games.

For a couple hours of straight gaming I didn't experience one crash, and my temps were around 65-70c, that's 10-15 degrees cooler.

 

This leaves me to believe the card is overheating while under stress, which doesn't make sense to me because the internet is telling me 80c is a normal temp while gaming.

 

Conclusion:

 

I have already contacted EVGA and they agreed to let me RMA the card, but I am hesitant to do so because I know I am just going to get a refurbished or a previously RMA'd card and I want to be absolutely sure it's defective before I send out my brand new card for replacement. Should I just replace it and hope that I won't get the same problems?

 

Is it possible maybe that the VRAM is overheating and not the GPU? I don't know but I need insight! Please help!

 

Thanks again for taking the time, The LTT community is awesome, you guys rock!

[Curufinwe_wins]

Download msi afterburner and watch the temps as you play games. What do they reach?

[failblox]

First of all, thank you for providing so much information. It helps a lot.

 

For starters, you might want to try other monitoring applications to make sure that the program you're currently using to read temps isn't faulty, as well as trying to run some benchmarks instead of games to see if you get the same results. Other than that, I can't think of much to do other than taking the card apart and replacing the thermal compound. While doing that, check the card to see if the card physically looks fine (i.e. checking if thermal pads are on the right places, fans are plugged in correctly, etc.). 

 

If none of that works then I'd go with the RMA from EVGA.

[nrc0]

Download msi afterburner and watch the temps as you play games. What do they reach?

 

Doing so now and I will report back after about an hour of gaming or when/if I crash.

 

Just so you know though I did mention that with precisionX I hit 80c on average during the most intensive gaming.

[Spectator]

Let us know what you find a screen shot would also be very helpful!

[nrc0]

So I played Rocket League for about a half hour then I got a crash to desktop. Here are the Afterburner results...

 

[nrc0]

First of all, thank you for providing so much information. It helps a lot.

 

For starters, you might want to try other monitoring applications to make sure that the program you're currently using to read temps isn't faulty, as well as trying to run some benchmarks instead of games to see if you get the same results. Other than that, I can't think of much to do other than taking the card apart and replacing the thermal compound. While doing that, check the card to see if the card physically looks fine (i.e. checking if thermal pads are on the right places, fans are plugged in correctly, etc.). 

 

If none of that works then I'd go with the RMA from EVGA.

 

I know how much of a pain it is not having enough info so I provided as much as I could thanks for the response!

 

I will try some benchmarks tomorrow for sure.

 

Doesn't taking apart the card void the warranty? I think i'd rather just send it in the go through the hassle of reapplying compound. Interesting thought though.

 

The card does indeed look fine at a glance but I noticed the card is extremely hot to the touch, like I can't hold my fingers on the same spot for more than a few second. But then again maybe 70-80c just feels that hot i'm not sure.

[mad dudy]

where did buy it from? if its amazon you can just return it for a new one not a refurbished one.

[nrc0]

where did buy it from? if its amazon you can just return it for a new one not a refurbished one.

I wish. Amazon offers 30 day returns and unfortunately I ordered it on Oct. 15 so too late for that

[nrc0]

The problem is I just don't even know if its the card at this point.

I think it has to be, but the other reason why I'm hesitant to send it back is because what if that's not the problem? Then I'm exchanging a perfectly good 980 for a refurbed one, waiting about 2 weeks to send it out and receive it, and STILL have the same problem.

Im coming to you guys for help because I'm between a rock and a hard place and I need some insight. Thanks again

[failblox]

I know how much of a pain it is not having enough info so I provided as much as I could thanks for the response!

 

I will try some benchmarks tomorrow for sure.

 

Doesn't taking apart the card void the warranty? I think i'd rather just send it in the go through the hassle of reapplying compound. Interesting thought though.

 

The card does indeed look fine at a glance but I noticed the card is extremely hot to the touch, like I can't hold my fingers on the same spot for more than a few second. But then again maybe 70-80c just feels that hot i'm not sure.

It seems like the card is the problem to me since when your 760 works fine in your system. Maybe it's power delivery? The 980 sucks up more power than a 760 so it's possible that could be the problem, but I doubt it since your PSU is pretty good.

 

Just quote me whenever you do the benchmarks so I get a notification

 

Yes, taking apart the card does void the warranty, but if you put it back together correctly they'll never notice. Same with overclocking and flashing the graphics BIOS - as long as you revert it before you send it in they won't be able to tell. But then again this assumes you can put it back together without much hassle. If you don't want to deal with this, then just send in the card.

 

It's perfectly normal for components to be too hot to touch. I used to do that with my graphics card...but I'd rather not get burned, so I stopped checking temps that way 

[Curufinwe_wins]

yea those temp and utilizations look totally normal to me. Honestly I don't know.

[Sonar]

Oh, I know...

if you have a small motherboard, like mITX, you may experience problems with south bridge. Your videocard (and the 980 is hot as hell) may overheat it. Check the temperature of the south bridge by Aida64 or smth. else. The temps should be around 60-80~85C.

The only solution, as I see (if you have the overheating of the bridge, of course) is the replacement of the motherboard. Try to avoid the small-format motherboards, they always have the problems.

[nrc0]

It seems like the card is the problem to me since when your 760 works fine in your system. Maybe it's power delivery? The 980 sucks up more power than a 760 so it's possible that could be the problem, but I doubt it since your PSU is pretty good.

 

Just quote me whenever you do the benchmarks so I get a notification

 

Yes, taking apart the card does void the warranty, but if you put it back together correctly they'll never notice. Same with overclocking and flashing the graphics BIOS - as long as you revert it before you send it in they won't be able to tell. But then again this assumes you can put it back together without much hassle. If you don't want to deal with this, then just send in the card.

 

It's perfectly normal for components to be too hot to touch. I used to do that with my graphics card...but I'd rather not get burned, so I stopped checking temps that way 

 

So I ran 3 benchmarks. Firestrike, Cinebench, and Unigine Heaven.

 

I ran multiple instances of each benchmark one after another for about 2 hours total and I am just baffled. NO crashes, not even once.

My GPU temp never exceeded 80c and my results were nothing out of the ordinary.

 

Even after all that testing, I started up Rainbow Six: siege, Rocket League and Assassin's creed Syndicate and the games are still crashing I just don't understand.

 

I set up my second monitor so I could play and monitor at the same time and I found out that while my main screen with the game locks and freezes, my second screen is displaying a BSOD. I did multiple trials and I experience 2 different kinds of BSOD's. One said "MEMORY_MANAGEMENT dxgmms2.sys" and the other said "SYSTEM_SERVICE_EXCEPTION" It looks like that is what is crashing my games.

 

I tried looking them up online but I can't find a straight answer to what's causing it and just wondering if you or anyone knows what this could be?

[nrc0]

Oh, I know...

if you have a small motherboard, like mITX, you may experience problems with south bridge. Your videocard (and the 980 is hot as hell) may overheat it. Check the temperature of the south bridge by Aida64 or smth. else. The temps should be around 60-80~85C.

The only solution, as I see (if you have the overheating of the bridge, of course) is the replacement of the motherboard. Try to avoid the small-format motherboards, they always have the problems.

 

I managed to get a screenshot of my temps about 2-3 mins before I got a crash, these numbers were pretty consistent while playing, does everything look good?

 

[failblox]

So I ran 3 benchmarks. Firestrike, Cinebench, and Unigine Heaven.

 

I ran multiple instances of each benchmark one after another for about 2 hours total and I am just baffled. NO crashes, not even once.

My GPU temp never exceeded 80c and my results were nothing out of the ordinary.

 

Even after all that testing, I started up Rainbow Six: siege, Rocket League and Assassin's creed Syndicate and the games are still crashing I just don't understand.

 

I set up my second monitor so I could play and monitor at the same time and I found out that while my main screen with the game locks and freezes, my second screen is displaying a BSOD. I did multiple trials and I experience 2 different kinds of BSOD's. One said "MEMORY_MANAGEMENT dxgmms2.sys" and the other said "SYSTEM_SERVICE_EXCEPTION" It looks like that is what is crashing my games.

 

I tried looking them up online but I can't find a straight answer to what's causing it and just wondering if you or anyone knows what this could be?

It seems possible that the BSODs might be linked to outdated versions of windows of drivers, so try updating Windows, as well as all your drivers, to the latest possible versions.

 

You might also want to try running memtest86 for 12-24 hours to see if there's anything wrong with your memory.

 

Tbh, your situation is pretty puzzling to me. I was wondering if it'd be possible for you to test your games in a fresh install of Windows. If you have a spare drive lying around, try putting Windows on that and running the same games. If they still crash, then it's definitely a hardware related issue, but if not, then the installation of windows you currently have is probably a little bit messed up.

[nrc0]

It seems possible that the BSODs might be linked to outdated versions of windows of drivers, so try updating Windows, as well as all your drivers, to the latest possible versions.

You might also want to try running memtest86 for 12-24 hours to see if there's anything wrong with your memory.

Tbh, your situation is pretty puzzling to me. I was wondering if it'd be possible for you to test your games in a fresh install of Windows. If you have a spare drive lying around, try putting Windows on that and running the same games. If they still crash, then it's definitely a hardware related issue, but if not, then the installation of windows you currently have is probably a little bit messed up.

So about a week ago, I was already having these issues for about a week already (I hope that makes sense lol) and I actually did a fresh install of windows 10. I reinstalled some games and I was still having issues.

Microsoft was not letting me activate my version of windows for some reason so I thought that something funky happend with the download. Turns out I download windows 10 pro instead of home which is why it wouldnt let me activate because I only have a key for home and not pro.

So I wiped my drives again and installed win 10 home this time and could finally activate it. Hoping my problem would be fixed, I downloaded a few game (again) and I still had crashes.

Even after 2 installs of windows and countless driver reinstalls it didn't fix the problem, which leads me to believe its a hardware problem.

Good point about updating windows though, I think it updates automatically but I will check when I get home from work.

As far as memtest goes, I have already gone through 2 passthrough of it and it passed both times but I didn't realize I should leave it on for that long so I mighy have another go at it.

After much convincing, a good friend of mine, who also has the exact same card as me (EVGA Gtx980 SC) is doing me a huge favor and we are swaping cards for the weekend. I think this will settle once and for all if its the card so i'm hoping for the best here.

If he starts crashing I know its the card and if I keep crashing with his working card, then there are other factors at play here.

I really appreciate all the help you've givin me so far and I will post back here with the results of the swap.

Thanks again!

[MrTiC]

Min GPU temp 59 ??? is that with the sidepanel on and at idle  ? if so remove the side panel cause that way to hot

[failblox]

So about a week ago, I was already having these issues for about a week already (I hope that makes sense lol) and I actually did a fresh install of windows 10. I reinstalled some games and I was still having issues.

Microsoft was not letting me activate my version of windows for some reason so I thought that something funky happend with the download. Turns out I download windows 10 pro instead of home which is why it wouldnt let me activate because I only have a key for home and not pro.

So I wiped my drives again and installed win 10 home this time and could finally activate it. Hoping my problem would be fixed, I downloaded a few game (again) and I still had crashes.

Even after 2 installs of windows and countless driver reinstalls it didn't fix the problem, which leads me to believe its a hardware problem.

Good point about updating windows though, I think it updates automatically but I will check when I get home from work.

As far as memtest goes, I have already gone through 2 passthrough of it and it passed both times but I didn't realize I should leave it on for that long so I mighy have another go at it.

After much convincing, a good friend of mine, who also has the exact same card as me (EVGA Gtx980 SC) is doing me a huge favor and we are swaping cards for the weekend. I think this will settle once and for all if its the card so i'm hoping for the best here.

If he starts crashing I know its the card and if I keep crashing with his working card, then there are other factors at play here.

I really appreciate all the help you've givin me so far and I will post back here with the results of the swap.

Thanks again!

Glad to help!

 

Another possibility, which I think is more unlikely, is that some of the soldering on the GPU is failing. A while back, my 6870 starting crashing and producing artifacts, and at the time I thought it was dead and just bought a 760. Last summer, I stumbled upon a post on the LTT forums with someone experiencing the exact same issue with the exact same card as me, and his solution was to reflow the GPU, which basically involves baking it in an oven at 385F for 8-9 minutes. This melts the solder just enough to have it rejoin in places where it possibly may have failed. For me, this fixed the issue with my 6870 and it works now, but it's not a permanent fix and for a card as new as your 980, I doubt that's the issue. To be clear - I do not recommend doing this with your card.

[nrc0]

Glad to help!

 

Another possibility, which I think is more unlikely, is that some of the soldering on the GPU is failing. A while back, my 6870 starting crashing and producing artifacts, and at the time I thought it was dead and just bought a 760. Last summer, I stumbled upon a post on the LTT forums with someone experiencing the exact same issue with the exact same card as me, and his solution was to reflow the GPU, which basically involves baking it in an oven at 385F for 8-9 minutes. This melts the solder just enough to have it rejoin in places where it possibly may have failed. For me, this fixed the issue with my 6870 and it works now, but it's not a permanent fix and for a card as new as your 980, I doubt that's the issue. To be clear - I do not recommend doing this with your card.

 

Min GPU temp 59 ??? is that with the sidepanel on and at idle  ? if so remove the side panel cause that way to hot

 

So I think something is overheating in my computer, what it is, i'm not sure.

 

If I leave the side panel off during the most intensive gaming I get average GPU temps of 60c and I NEVER CRASH. But when I leave it on it heats up to 80c average, my computer sounds like a jet engine, and the game/entire PC eventually crashes. 

 

I used to play 1080p ultra settings with my old Gtx760 and never had any problems, my computer ran smooth for 18+months. I'm trying to now run 1440p ultra settings with the 980, which should and does handle it, but then the problems stated above happens. 

 

Sometimes I get bluescreens, sometimes the game crashes to desktop and I can send crash report to developer, or somethings I get the error, "The display driver has stopped working and has recovered." 

 

Any insight? Should I just RMA the 980 and hope for the best? God this sucks lol

[nrc0]

By the way I swapped cards with my buddy and either of us had any problems, granted he has a bigger case and more airflow so that's probably why, but I had no problems with his card in my system. I think I'm just going to RMA it I don't see another option

.

[iLenovo]

 

Hello and thanks for clicking my post!

 

Case - Bitfenix Prodigy

 

Motherboard - Asus Maximus VI Impact

 

CPU - Intel Haswell i5 4670k

 

GPU - EVGA GTX 980 SC

 

Memory - G.Skill Ripjaws X Series 8Gb (2x4gb)

 

Storage - x2 Samsung 850 SSDs (250gb)

 

CPU Cooler - Corsair H60 Liquid Cooler

 

PSU - Corsair HX650 80 Plus Gold

 

OS - Windows 8.1 OEM

 

Some Background...

 

So about a month ago I bought a new EVGA GTX980. It replaced my old Gigabyte GTX760. Note that I had the 760 in my PC for over a year and never experienced any problems.

 

Everything was going great for about a week, then I started running into problems... More specifically my computer crashes while playing games, and only games. I never have a problem when I'm not in a game. Sometimes it crashes to desktop, sometimes it locks my computer with a loud, repetitive audio lock which requires a full computer restart. Below is a list of games that I have experienced multiple crashes on, usually within 15 mins of playing, which render the games unplayable:

 

- Assassin's creed: Unity

- Assassin's creed: Syndicate

- Rocket League

- Rainbow six: Siege (Beta and Full Game)

- Sid Meyer's Civilization 5

- Fallout 4

- Star Wars: The Old Republic

- Counter Strike: Global Offensive

 

The only game I have not experienced this in is League of Legends, which is not a very demanding game in general.

 

Things I have done:

 

- Everything in my system is running default (No overclocks on CPU or GPU)

 

- I set up my second monitor for testing purposes (I only ever have 1 monitor connected for general use and while gaming) and I started up multiple games.

          I had benchmark programs running on the second screen, including EGVA's precision X, RealTemp, GPU-Z, and CPUID Hw Monitor Pro, as well as                       windows 10 task manager to watch memory and cpu usage. All the numbers were running within comfortable specs, the GPU never exceeded temp 80c

          and my cpu was staying around a cool 60-70c. I did multiple trials and everything was normal, but to no avail, I always got a crash 10-30 mins into playing.

 

- I uninstalled Nvidia drivers and installed current ones 2 or 3 times

 

- I updated my motherboard bios, as well as loading defaults on my entire bios

 

- I even reinstalled windows 10 (Saw this on a forum but still didn't work)

 

- I put back in my old Graphics card (Gtx760) and played a few games. Didn't experience one crash

 

- The only thing I changed when I started experiencing these crashes was my graphics card, I didn't change out any other components other than my monitor (but that wouldn't matter right?...RIGHT??)

 

- All my fans in my case are spinning

 

- I have tried both 1080p and 1440p resolution in games with same results

 

- I have not experienced a crash outside of playing a game

 

- None of my games are modded

 

- I took my card out of the case, de-dustified the whole computer, and reseated the card

 

Important Info:

 

Because I have a mini-ITX case, my Graphics card sits right up against my side panel. I decided to do a test.  

 

I took off my side panel, exposing the card and its 2 fans to fresh air, and I played some games.

For a couple hours of straight gaming I didn't experience one crash, and my temps were around 65-70c, that's 10-15 degrees cooler.

 

This leaves me to believe the card is overheating while under stress, which doesn't make sense to me because the internet is telling me 80c is a normal temp while gaming.

 

Conclusion:

 

I have already contacted EVGA and they agreed to let me RMA the card, but I am hesitant to do so because I know I am just going to get a refurbished or a previously RMA'd card and I want to be absolutely sure it's defective before I send out my brand new card for replacement. Should I just replace it and hope that I won't get the same problems?

 

Is it possible maybe that the VRAM is overheating and not the GPU? I don't know but I need insight! Please help!

 

Thanks again for taking the time, The LTT community is awesome, you guys rock!

 

What are your clock settings? If so, could you tell me your fan settings as you ramp up your clock settings? 

EVGA cards tend to run hot, just like mine, but a complete restart to use your computer sounds wrong. Until I know your clock settings I can help you much more.

[failblox]

So I think something is overheating in my computer, what it is, i'm not sure.

 

If I leave the side panel off during the most intensive gaming I get average GPU temps of 60c and I NEVER CRASH. But when I leave it on it heats up to 80c average, my computer sounds like a jet engine, and the game/entire PC eventually crashes. 

 

I used to play 1080p ultra settings with my old Gtx760 and never had any problems, my computer ran smooth for 18+months. I'm trying to now run 1440p ultra settings with the 980, which should and does handle it, but then the problems stated above happens. 

 

Sometimes I get bluescreens, sometimes the game crashes to desktop and I can send crash report to developer, or somethings I get the error, "The display driver has stopped working and has recovered." 

 

Any insight? Should I just RMA the 980 and hope for the best? God this sucks lol

Could you send me a log of your GPU under load using GPU-Z instead of AIDA 64 for me to take a look at? I find GPU-Z has more data to look at than AIDA 64 and that might help me to further figure out why your PC keeps crashing. It shouldn't be crashing at 80C, and that's a perfectly fine temperature to run at under load.

 

I agree with what @iLenovo said. We need some more information like clock speeds, fan speeds (% and RPM), etc. to get a better picture of what's going on.

 

There's a checkbox at the bottom of the GPU-Z "Sensors" tab where you can choose to log all the data to a file.

[iLenovo]

From some of the data I received, its not your computer, its your graphics settings. Try turning down your core clock and ramp fans 75%. 

[iLenovo]

It may be a motherboard/CPU malfunction, once again, make sure you have no malware, and if you can try to lube and de-dust your fans.