Hello guys,
 
i own this card since 28.11.16. Until yesterday it runned perfectly. 2 weeks ago i buyed a new 1440p display and i went with some good results on mid graphics with this card.
 
Then yesterday evening i could see artifacts on my monitor from one moment to another. It looked similar to this (not my photo since it came up only yesterday evening for 1 time now):
 

 
************************************************************************************************************************
Update 9/2/2019: just 10 minutes after i pushed the RMA forward i could take a photo of my V-Ram failing
 

 
****************************************************************************************************************************
 
After a restart this was gone. But i now have the problem that all the games are crashing with the same graphical settings they worked with for 2 weeks without a problem. Even when i lower the settings further it crashes now constantly.
 
- i did never overclock the card
- i installed the thermal mod kit from evga some weeks after i buyed the card when i was available. did this exactly like it was written down with no problem in 2 years.
 
After i tried lower settings i removed the card to check it out and clean it to see if its better then. Well the card itself wasnt really dusty since i cleared it from time to time:
 

 
But what i saw next shocked me:
 

 

 
 
The was some fluid like oily stuff coming right from the inner backplate to the PCIe connector! I dont usw watercooling and nothing could have happened with water or something. That i was sure about. So i decided to take a look behind the backplate. And what i saw next shocked me even more:
 

 

 
It looks like the cooling pads just melted away.... im not sure it thats what is it. Maybe anyone of you guys have a idea?
 
I can only guess how it looks in the inner card itself with thoose molten cooling pads...
 
 
The artifacts from the picture above didnt come back but the card still crashes in games with other artifacts. So should i RMA this card? Warranty runs until 2/7/2027 so that wouldnt be a problem i guess.
 
Greetings
 
Kabs
 
 
 

Cool GTX
Silicone - does leak - squeeze out - of the Thermal Pads = Normal after thermal cycling
 
Game Crashing:
 
1) check for known bugs in Game
 
2) Unstable MB OC
 
3) Driver Update - Newer Nvidia driver is NOT always better driver

As i said. It runned perfectly until yesterday thoose artifacts appeared. I did not change anything so i dont think any of that is the problem.
 
Thoose artifacts i only saw on busted cards with defective DDR and such.

Cool GTX
What Monitor & are you OC the Monitor ?
 
What have you changed in Nvidia Control Panel ?
 
What OS & have you installed any Updates ?

Brand new LG 34GK950G and i did not oc from 100 to 120 Hz since it doenst make sence with that GPU atm.
 
Did not chance anything in control panel. Display runs 3440 x 1440 100 Hz native.
 
Win 10 all updates 1903 (build 18362.295)
 
 

Update2:
 
sooo... it comes back if i play 3440 x 1440 p every time... sometimes it needs a bit longer, sometimes not...
 
So it first came up after theese artifacts. before it runned completely ok for 1-2 weeks. and now its only happen on uwqhd and not on FullHD.
 
 
Conclusion: This cant be a software problem since everything worked as mentioned above and i didnt change anything before the problem happend first.
Really seems to be a somewhat busted card. Maybe it gets to hot when using UWQHD.. maybe one hardware part DDR or such is defective.. not sure. But this isnt supposed to happen.
 
I will test a bit further if i am 100% sure the above is correct. If it is i should RMA? correct?
 
Best Regards!
 
Kabs

Cool GTX
Could be a bad video cable
 
Could be a monitor issue

Monitor and Cable brand new. All went fine until the artifacts, wich definatley came from the GPU. When it crashes games to windows i can see everything, the game just hangs up and crashes. GPU performance in task manger = 0.
 
I can look Videos on new monitor for hours without a problem at the moment. But if i play games and the GPU needs to work harder, it crashes the games.
 
i played for 2 Hours on Full HD without a problem. On UWQHD games crash every 10-15 minutes. Tested with Shadow of the Tomb Raider, Ancestors Humankind Odysse and 2 other games.
 
So its definately not a monitor or cable problem.

Hi,
 
i dont have another pc here to test that :L
 
Here a screen of task manager when it happens.
 

 
i even checked if i have not set powerplan to performace. Still happens only on the UWQHD.

Sajin
You can have the card tested at a local pc shop.

you think they would? but they even would need my monitor? i am not so sure they would do this if i havent buyed the card there
 
Screen when game is still running:
 

 
 
 
Update:
 
i now had precision and task manager on my other screen running to see what happens on their end when the problem comes in.
 
Conclusion: game crashes and only noticable thing is = 1 second after crash task manager shows 100% gpu usage for a milisecond and then it goes directly to 0% as shown in my first screenshot.

Thanks for the great support here btw! coming from german forums as it seems dead..
 
sadly i think i tested everything that could be a problem other than the gpu itself...
 
testing the gpu and display on other system is really a problem. the friend i mentioned above lives too far away sadly. i will try to check if i get to test that somewhere near but it seems very hard to do so.
 
 

jonkrmr
If you are getting the "Space Invaders" artifacts, this would indicate the ram on the card is failing.
 
The fact it is not doing it at 1080p resolution but at 1440p resolution could be that your card has 8GB ram on it and at 1080p resolution not all of your ram on the card is being addressed and used.
At 1440p resolution more of the 8GB ram is addressed and used. The bad ram could be higher in the ram allocation on the card.
 
Basically, you might have a single bad memory chip on the card that is higher up in the ram allocation on the card and the problem only shows up at higher resolutions where more ram is addressed and used hitting the "bad" ram chip and causing the artifacts.
 
Either way, the "space invaders" artifacting is a indicator of failing ram on the card and the card should be RMA'd if it is still under warranty.

Hi,
 
i now have tested further with 2560x1440p and UW 1080p. On the 2560x1440p it crashes really often and now i also have 1 crash with UW 1080p at extreme graphics.
 
You could be really right with the above as i can see that the crashes come more often if the used graphics power rises. The crash at UW 1080p extreme graphics took like 10-13 minutes while 2560x1440p took only 7-9 Minutes. the crashes with 3440x1440p are even more often.
 
Just now i got a crash with UW 1080p mid graphics after 15-20 minutes...
 
 
Well i guess youre right with the higher ram allocation! It looks more and more like it!
 
But i dont have seen the same artifacts from 2 days ago again. a friend told me it can still be the ram and not nessecary to show the artifacts again. is that true?
 
Think its time to RMA then :( The fact alone that it all was good before the artifacts showed up 2 days ago tells me that youre right and a higher allocated ram burned down.
 
Will do a bit more testing with another game but RMA seems unavoidable..
 
 
 
 

opened a support ticket but didnt got a email to confirm until now :L

I did check if there is something going on with my PSU voltages and other measurements:
 
voltage idle:
 

 
all measurements 2560x1440p ultra graphics:
 

 
 
all measurements 3440x1440p ultra graphics:
 

 
 
 
All seem pretty fine arent they? only thing wondering about is the gpu usage from tomb raider in task manager.
 
Still the games crash.

looks like RMA time ..

Once a card starts artifacting & crashing it will continue to do so and get worse as time goes on until one day the card will no longer work at all.
 
Chip failures (GPU - MEM - VRM) do not magically fix themselves
 
Even if you happen to find a resolution and graphic settings that don't exhibit the problem with the card does not mean the problem has gone away.
 
Eventually, the card will get stressed enough or the bad memory addressed and BAM! problem rears its ugly head again because it was always there..........
 
I will say a prayer for you that your RMA replacement card will be a good chip and overclock better than the one you are sending back