1080TI Over Heating 1.5 years in - RMA?

[ryan.o.kean]

Specs:
MSI z270 Motherboard
i7 7700k
16gb DDR4 Corsair ram
evga 650 80+ gold power supply
1440p Gsync Dell monitor
Gigabyte GTX 1080TI OC 11G

Issue:
My 1080ti is quickly jumping to +90* C temperatures upon any sort of graphical stress. I got a dual monitor setup about 2 weeks ago, and was attributing some stuttering and fps issues to that. I have since removed the other monitor and narrowed down the issue to what i believe to be my gpu hardware. Case is very open, no dust, 3 case fans, 212 evo on my cpu. Removing side panel barely did anything.

This first came to my attention as my monitor started to randomly lose signal a few days ago. I would still have audio and power to the rest of my system. I would also hear a windows notification similar to when something is unplugged. I've had this exact system for about a year and a half, and haven't updated any software or anything recently.

I noticed my temps were jumping almost instantly up to ~94 max when running games. Next would come fps drops and eventually lose of monitor signal.

My first step was to DDU my drivers and clean install. This completely fixed my issue for ~24 hours under heavy stress. Note: I did this before i noticed my temperatures, so it is possible my temps were still high, but my fps was perfect and didn't lose monitor signal. I checked one of windows logs and noticed a "kernal 141 error" I can get more specific info once I'm off work.

Last night my card was constantly 85-94* with awful fps throttling. I replaced the thermal paste on my card, and this helped slightly. See imgur link for screenshot of my temps with new thermal paste and fans at 100%
https://imgur.com/gallery/XjEZjbL

I feel this is still too high to be normal, especially if my system has performed fine for the last 1.5 years. What would you guys recommend as the next step? RMA the card?

I really appreciate any input or support, thanks!

 

[king3pj]

I think Gigabyte has a 3 year warranty on GPUs so if you can get them to send you a new 1080 Ti that's what I would do.

I don't have a 1080 Ti but I've read that temps in the low 80s are normal with that card. 85 is a little on the high end of that but if it was maxing out there and you weren't having other problems I would say it's probably okay. Going up into the 90s, losing signal, and getting kernal errors is not normal behavior though.

Also, a second monitor should not cause any problems for a 1080 Ti. I've been running two 1440p monitors off a 1070 for a couple years now and SLI 970s before that without any issues. Hell, we run dual monitors off the integrated graphics in a i5-4460 on all the machines in our office.

 

[arif1752002m4a1]

Clean you graphics card with compressed air.
If temp still rise.
Reapply thermal paste with new one. (good quality thermal paste)

Some times power supply also cause system temp rise without load. Like CPU GPU even Hard disks too. So check your system with any good PSU.

 

[kerberos_20]

if its under warranty, then u shouldnt reapply thermal paste
other pc temperatures are ok?
too high temperature inside PC will not help it, GPU is takin fresh "cold" air from inside your PC, make sure your intake fans are working and there anrent too many obstacles (dust, cables)

 

[ryan.o.kean]

- It is still under warranty. I will likely start the RMA process, but I was hoping for some more experienced people's opinions before doing so. I've heard mixed things about thermal paste voiding the warranty, however I don't think they'll have any way of knowing that I did

-I was under the impression the second monitor was causing stuttering due to different refresh rates/resolutions, but i think it was just coincidental timing It does seem to be a common problem for people, notably while gaming with one monitor and using YouTube on the other.

- I did thoroughly clean my card with compressed air. Case is dust free and very open with decent cable management.

I wouldn't be surprised if it was capping at 85*, but not with fans blasting at 100% and still jumping over. Doesn't seem quite right.

Any other input would be much appreciated! Thanks everyone.

 

[Danra]

What is your room temperature?

In your last post you said, "Case is ... very open with decent cable management."; does that simply mean the case is closed and there is a lot of room for air movement inside the case? In your first post you mentioned opening the case did little, however, it did something, better cooling or worse?

I suggest you open the case, use a box floor fan set at an angle of about 60 degrees forcing air from front and out of the back of the case. If that helps more than simply opening the case, then you have the wrong case/fan setup.

Note that Gigabyte is widely known for poor cooling compared to other manufacture cards. On a side note, even within other company cards some coolers are better, so researching prior to purchase is advisable.

Also, the fans on the Gigabyte are not the best and are known to die earlier that those from other companies. Have you been looking at fans RPM since the card was new to compare current RPM?

There are some known issues with your card cooling, however, in this instance there are unknowns because you did not list them. What is your room temperature, exactly what case do you have, how many fans are in it, what direction are each of the fans moving air and what is the location of each of those fans, including the fan for your CPU water cooler?

Exactly what power supply do you have and how is it oriented, fan pulling air from inside the case or fan pulling air into the PSU directly from outside the case?

Do you have monitoring software, if so what is/are it's name?

There are many other questions I can supply, however, at this time this is enough. If I had my hands on your computer I could make determinations quickly, however, we obviously have to do this at a distance and that requires time for the best diagnosis.

Note: I did not proof read this, so, if there are grammar errors please accept my apology.