Hi guys,
So here's the deal. My faithful 1080 FTW gave up on me (after 3 years and 9 months of service) early in the morning yesteday by essentially self-combusting (almost) while idling. Was smelly enough to wake me and my GF up in the bedroom while the PC was in the living room lol.
 
 
Anyhow I got over that quickly and got on the local marketplace for used stuff online (I live in Bulgaria). Got a good deal on a 1080Ti FTW3 still in warranty and brought it home today. I installed it and ran 3Dmark (9546 score),Heaven and Superposition. The temperature seemed fine, the boost clock as well. I also played World of Warcraft at Ultra, Max AA 1440p for about 40 mins. It had no issue whatsoever and substantially higher fps than what I had with my now dead 1080 FTW.
 
However, when working remotely today after installing the card (using Remote Desktop), the following line appeared:
 

 
It changes colours when I click next to it and is only visible in the remote desktop window. I had to take a photo with my phone cause when I do a print screen there's nothing there:

 
So I thought it could be a Remote Desktop issue (even though I had been using the program ever since we got to stay home here more that 40 days ago and it had never happened before). But of course it kept bugging me, so I got a few more benchmarks. 
This is a screen from the GPU benchmark of UserBenchmark

 
 

You can see the same line there, also chaning colors during the benchmark. And the benchmark actually fails in the end:

Then I tried FurMark - the 1440 preset and the 1080 preset. No issues, lines, etc. Max temp was 70 degrees. This was the score for 1080:
 
 
 
 
So... I'm trying to figure out if there is an issue with the card or not. :-D Any ideas? 
 
Thanks!
 
Probably the rest of my config could be useful:
i7 8700K @ 5Ghz
Asus Maximus X Hero mobo
16 GB G.Skill Trident Z @ 3200mhz
Seasonic Platinum 860W

bob16314
That looks like you have "Show FCAT overlay" turned on in the EVGA Precision settings..Check to be sure it's not on.

You are 100% correct dude! That resolved the issue. Thanks a million! :-) Phew.. I am always super nervous when buying used stuff and prolly that is why I very rarely do. I was just waiting for Ampere and still am and figured I did not want to splurge on a 2080 or something cause that would mean skipping Ampere. So getting the 1080Ti for around EUR 390 seemed like a good deal. 

bob16314
Yep, figured..Glad that was all it was
 
Welcome to the forums, by the way..Be sure to read  if you haven't yet..Lots of good info there for you.
 
Waiting on the Ampere release myself..Way overdue to upgrade from my 760 SC.

 
Thanks! I will most definitely check it out. Btw I could not help noticing we had a pretty similar configs. I also delidded my 8700K and used Conductonaut :-D It was probably a worse idea in general than buying a used 1080Ti considering the risk and the voided warranty, lol. But it is still working at less than 50 degrees under load so I guess it was worth it. :-D

What monitor(s) are you using & what specs are your set to use on them ?
 
What is your Room Temp?
 
What are your case fans set to RPM or % ?
 
What case are you using ?
 
When you compare to others, you need to duplicate their setup & environment ....aka ... its all in the details

Is your Kraken x62 fans pulling air into the case or pushing air out?
 
Try taking the side cover off and see if it has an effect on your idle temps.
 
Also would not hurt to re-tim it and check the condition of the thermal pads and replace those if needed.
 
 
 

If your case fans are controlled to your CPU temp .... the AIO will keep temps down .. your case fans will run slower
 
Test All fans 100%
 
Best to have more fans blowing in to the case than out of the case - positive pressure is good for cooling

Hi again guys. Well I opened up the card and changed the thermal pads and paste ( as bob16314 suggested). I used the Arctic thermal pads 1.0mm and 1.5mm, and the Arctic MX4 paste. 
 
Here's what the card looked like after two years of mining:

 

 
 

 

 

 

 

 

 

 

Thankfully it survived the intervention (just one photo with the new pads but I've put them everywhere they needed to go :) ), however, there was no difference in the temeperature
 
Here's the idle before:

 
And here's after:
 

 
Granted the first one was with taken with ambient temp 22C and the 2nd one with ambient temp 25C. The core also seems a bit hotter now under load. Went up to 73C in Heaven and  I'm pretty sure I put the correct amount of paste but now I'm tempted to open it up again and see if it was enough/too much. :D 
 
A non-related issue I've been noticing the last couple of days. I get flickering from time to time on both monitors. This did not happen on my 1080 FTW. Turned off Nvidia Multidisplay Power Saver - was still there. It does not happen too often but is still a bit annoying. I read a bunch of threads that at some point this was due to the Nvidia drivers but it seems they fixed them at some point.

arestavo
With non-conductive thermal paste, you cannot put too much on.
 
No, really, you can't. Unless you're deliberately trying to cause issues with pounds of the stuff.
 
You can put too little on and cause problems since the die doesn't have an IHS. So, if you do open it back up again and see that the paste spreading is missing coverage in spots (likely the corners) - you've put too little on. I favor the X pattern when putting paste onto the GPU - hasn't failed me yet!

Thanks for the reply. I always stress over the amount I put on my CPU but at least it is a lot easier to check if something went wrong and reapply. Here it is a bit more work. :D I also used the X pattern to apply the paste. It was a bit "skinny" X truth be told.