d0zer44
still keeps crashing in debug mode. 
 
it seems it crashes faster in league of legends than.
 
15 minutes stress test with furmark  on 70° temperature on gpu doesnt crash my system. but 5 minutes in league of legends with 40° temperature crashes my system?
what do you guys think of that weird behaviour ?

This is an odd situation where the system is stable with a brutal stress test like Furmark but not in gaming.  Usually this is going to indicate a corruption in the game but you are having issues with multiple games.  With this set of issues, I would suggest either testing the card in another system or contacting EVGA to possibly get the card replaced. 
http://www.evga.com/about/contactus/

d0zer44
still keeps crashing in debug mode. 
 
it seems it crashes faster in league of legends than.
 
15 minutes stress test with furmark  on 70° temperature on gpu doesnt crash my system. but 5 minutes in league of legends with 40° temperature crashes my system?
what do you guys think of that weird behaviour ?

Sounds like the card is faulty to me. The reason I say that is because furmark will cause the card to run at lower voltage because it's blacklisted by the nvidia driver. When running games the card will use full voltage unlike when running furmark. How old is your psu? Does manually underclocking the card with msi afterburner give different results than running in debug mode?

Sajin

d0zer44
still keeps crashing in debug mode. 
 
it seems it crashes faster in league of legends than.
 
15 minutes stress test with furmark  on 70° temperature on gpu doesnt crash my system. but 5 minutes in league of legends with 40° temperature crashes my system?
what do you guys think of that weird behaviour ?

Sounds like the card is faulty to me. The reason I say that is because furmark will cause the card to run at lower voltage because it's blacklisted by the nvidia driver. When running games the card will use full voltage unlike when running furmark. How old is your psu? Does manually underclocking the card with msi afterburner give different results than running in debug mode?

 
so i just have to underclock the memory and core clock by for example  10% or to the reference clocks by nvidia ?
do i need touching the shader clock?
 

I'm not sure if this will help or not, but I was getting black screens where the sound continues while playing Skyrim Special Edition before I switched the power mode from the new drivers default of Optimal Power back over to the older driver series default of Adaptive.  Prefer maximum performance would probably work as well, but would use more energy as the GPU would spool up even for the smallest task.
 
My 1080 also runs very cool under an AIO cooler - 39C or under, even overclocked and at 4K res.
 
For me my computer goes to a black screen, with the sound continuing to play, after a random period of time - usually longer than a half hour, and sometimes as much as two hours. Control Alt Delete doesn't do anything, unless I unplug the HDMI cable and plug it back in - then I could see the Control Alt Delete screen. The same trick doesn't bring the game back when I try doing that instead of Control Alt Delete. Doing the below steps seems to have fixed the black screen issue for me, so I can only guess that it has something to do with the cards very advanced power states and them changing during gameplay.
 
 
1. Open up the Nvidia Control Panel (NVCP)
2. Under 3D Settings on the left side, select "Manage 3D Settings"
3. Under Power Management Mode, change it from "Optimal Power" over to "Adaptive" (which used to be the default with older drivers)
4. On the bottom right of the NVCP click the box marked "Apply"
5. Play your games and hopefully without those random danged black screens!
 
 
For reference, I'm running Win10 Pro x64 AE, and a 1080 ACX 3.0 with 375.70 drivers. And unfortunately I don't play Forza 3 or LoL, but I've put in God knows how many hours of Skyrim SE in the past week.

d0zer44

Sajin

d0zer44
still keeps crashing in debug mode. 
 
it seems it crashes faster in league of legends than.
 
15 minutes stress test with furmark  on 70° temperature on gpu doesnt crash my system. but 5 minutes in league of legends with 40° temperature crashes my system?
what do you guys think of that weird behaviour ?

Sounds like the card is faulty to me. The reason I say that is because furmark will cause the card to run at lower voltage because it's blacklisted by the nvidia driver. When running games the card will use full voltage unlike when running furmark. How old is your psu? Does manually underclocking the card with msi afterburner give different results than running in debug mode?

 
so i just have to underclock the memory and core clock by for example  10% or to the reference clocks by nvidia ?
do i need touching the shader clock?
 

Underclock the memory & core clock as far as you can. Shader clock shouldn't even be an option inside msi afterburner.

i have now underclocked by 10% and still getting that weird black screens.
i also changed the power management mode in nvidia control panel to adaptive.
 
i have now paid with paypal for the advanced rma and am now waiting for my replacement card as i dont know what else to do.

Let us know how the replacement works.

Replacement is here since 2 days. Now it doesnt lose more signal to monitor but the card crashed twice with nvlddmkm.sys error. 
So i guess the driver from nvidia are causing the crashes.
 
Then i found the solution for those crashes for my config. I autoupdated in Hardware Manager everything under "System Devices". Intel Management Enginge found Updates for 3 devices and since then, i could play 4 hours without a single crash.
 
 

Glad your problem is now resolved.