landerx74

tekanna
I got my first black screen+100% fan speed today when playing rocket league... ROCKET LEAGUE! Not the most demanding game. Albeit I had the card overclocked +75mzh on the core and +175mhz on the vram. Full power and everything. I have dialed down the overclock a bit now and tested with heaven and it has not crashed again. It would be nice if you could overclock the card without being stressed out about this crash. We'll see what the future holds I guess...
 
Would be nice, like somebody else mentioned, if EVGA could replicate the crash in their labs and see if there's a solution to the problem. 

If I was you I would return the card to the place you purchased it and see if you can get a store credit. Get a 1080 that does not have "FTW" in the name. I hope that EVGA now sees this as a big enough problem to issue a BIOS update/work in progress report etc.etc. something to tell the scores of users that this problem is being addressed and that a fix is coming soon.

I have emailed my retailer. I'll probably get an answer on monday or tuesday. I'm considering the ASUS strix or the Gigabyte G1 instead. A bit more pricey in my country but maybe it's worth it if they don't crash?

OK, I did not read 6 pages of post .....
 
My best guess - heat related crashing - (thermal induced expansion)
 
Has anyone with this issue -- pulled their cooler and checked the thermal pads & TIM ?
 
10X loop and inspect the solder joints ?

HeroicArchetype
I am running in debug mode and setted the fan curve with different settings. Didn't crash, could be for the debug mod tho, I will try without tomorrow.
 
Anyway temps should not be the issue since people got this crash even at 60°.

Thermal induced (not max temp)  something changes as it heats / expands.
 
Solder joints, PCB

I got a video of the black screening as it happens, for people to see what it looks like. I'm on mobile and I used the Upload file to the server option. Not sure if that worked or not

gelectrode

Cool GTX

HeroicArchetype
I am running in debug mode and setted the fan curve with different settings. Didn't crash, could be for the debug mod tho, I will try without tomorrow.
 
Anyway temps should not be the issue since people got this crash even at 60°.

Thermal induced (not max temp)  something changes as it heats / expands.
 
Solder joints, PCB

 
That was my first guess as well. I was gonna check my self but i dont have any spare thermal pads around or paste and didnt want to spend more money (iam broke and jobless atm) on ordering it in from the states for the pads. I thought maybe pads to short or missing off of the power components or even the ram. My problem has not yet returned since removing my overclock as well as reseating my power connections in my psu and gpu (even though they were snug on both ends).

Good to see your issue is resolved

tekanna
A small update.
 
I haven't had a crash since I dialed back my overclock. Now I only have power limit and thermal limit maxed out, nothing else. So far so good. I believe that overclocking might be the problem for the majority of the cards due to the high factory overclock. 

Removing Your OC or the factory OC ?

most of us are getting the crash with just the stock overclock and not overclocking the card ourselves.

Cool GTX

tekanna

Cool GTX

tekanna
A small update.
 
I haven't had a crash since I dialed back my overclock. Now I only have power limit and thermal limit maxed out, nothing else. So far so good. I believe that overclocking might be the problem for the majority of the cards due to the high factory overclock. 

Removing Your OC or the factory OC ?

Sry. English is not my first language.
 
My own OC is removed. I'm dialing in a new one as we speak. I have got it working with no problems by just upping the power limit and thermal limit to 120% and 92C, leave everything else at stock.
 
The OC I'm working on now is:
power limit and thermal limit - 120% and 92C with priority on the temp.
+75mhz on the core, it shows 2088mhz in heaven benchmark
+450-500mhz on the vram, shows 5405mhz in heaven benchmark
Without touching voltages
 
I think that was the problem before. I upped the voltage to max and got the crash. I was naive enough to think that more voltage would automatically yield higher boost/clock speeds and i pushed it too far.

Yes, voltage is the last thing to adjust - not the first.  Heat impacts the 10 series & voltage = Heat

Yes you are right.
 
I seem to have a stable overclock for now. It sucks that people are getting this crash with stock settings. I will update if anything else happens.

ZBeaverEater
Hey guy's, I re- seated my card, re-installed my drivers and I'm still getting strange white lines across my screen whenever I do a cold boot. When the system warms up, they go away. I decided to try debug mode today, and when I did, The top one third of my monitor displayed little color full squares and rectangles. I rebooted, and all was good again. No crash today, but things aren't quite right. What are your opinions? RMA? I paid over $700 with tax for this card.

Sounds like you got yourself a defective card. Have you tried a different display cable/output?

thenson88
I have had this issue since launch and the only solution for my 1080 FTW has been to uninstall Precision X OC. I can run the same clocks with MSI Afterburner and if it gets unstable it will just do a driver crash instead of the black screen and fan revv.

Wow so you're saying since you uninstalled Precision x OC you never had a black screen with your fans revving up? So just a driver crash?? And how long has it been since you uninstalled Precision x OC?

Seems to be a lot of confusion with how GPU boost 3.0 works and what the clock speed will be from the factory. On the 10 series the boost clocks on a cards specs sheet is only the min working boost clock from the factory, it can boost higher and almost all of them do. They also can go below the boost to the base clock when throttling and you see this happening with factory settings FE cards all the time due to thermals. This is all normal and how is dynamic clock speed with boost 3.0 work, yes it can be a pain in the rear if you want a stable clock. 

Nyy82
So, after days of troubleshooting I got this stupid **** figured out. It's a power management issue with in Nvidia's Software.
Right Click on the Desktop, Go to Nvidia Control Panel, Go to Mange 3D Settings, go to power management mode and change it to "prefer maximum performance". In Optimal, it's actively adapting the power and I guess it hiccups, it under powers the card and the card activates it's fail safe. Which is black screen, fans 100%.
 
2 days now, and the card is stable. This is over the course of 3 hours of SimCity, and 2 hours of BF4. Both games running @ 4k, ultra. 

Strange. I don't own any 1080's. My cards (4x maxwell titan x) work perfectly fine in games running driver 372.54 using default driver settings.

I believe removing the heatsink or tampering with the cooling apparatus will void your warranty. Quite clearly hardware related defect which may be why EVGA is silent because they have to conduct thorough testing before issuing recalls or anything drastic. Best bet is to just RMA... I'm still waiting for RMA approval.

One thing an EVGA tech recommended was to check for 12volt rail voltage stability under load to eliminate faulty or unstable PSU. Your mobo bios can give you readings at idle but you'd need AIDA or other software to give you the readings at load.

Per tech: The 12v rail should read between 11.4v and 12.6v (+/- 5% of the rail's rating). It should be stable, with little to no fluctuations. A fluctuation higher than .1v, or more often than every 10 seconds could be an indication of an issue with the 12v rail

pwner1984
I believe removing the heatsink or tampering with the cooling apparatus will void your warranty...

No it will not as long as you don't physically damage anything and return it to stock when you RMA: http://www.evga.com/suppo..arranty/graphics-cards/

I'd just like a response that it's being looked into before I return this and take the 15% hit. :\