#1 Open the nvidia control panel, under manage 3d settings select the global settings tab, change power management mode to prefer maximum performance, reboot pc. 

#2 Run GPU fans at maximum speed while playing your games.

1. Download and install MSI Afterburner.
2. Launch afterburner and set the fan speed (%) slider all the way to max.
3. Click apply to bring changes you've made into effect. GPU fan should get louder at this point.
4. Test games. Leave afterburner open while testing.

#3 Report back if you issue is resolved or not.

Next step would be to try increasing the voltage inside afterburner, I would set it to max just for testing, make sure your fan is also running at max when doing this. If increasing the voltage doesn't help try downclocking the card on the core & memory by 105MHz to see if it helps stop the black screen. All 970/980's seems to be hit or miss on this issue. I haven't seen very many people with classifieds reporting the issue though.

Raising Voltage to max (+37) did nothing from what i can tell. Same voltages as with 0. The downclock however, worked. Could finish Sky Diver Demo without a blackscreen. Thanks. 
 
So just a bad factory overclock. Probably wouldn't be a problem if i could just raise the voltage a bit past 1.212 though.
 
Edit: Would a bios update for the card help here? 

A vbios update may help, contact evga and have them send you the latest vbios for your card to see if it helps resolve your issue or not.

Got the bios but it didn't help. I will use it downclocked until end of next week when i have money for cross RMA and send it in then. 

I remember way back when.... when my new rig was new last September hearing of people having issues running Skydiver back then. Me too. Black screens.. I ended up buying 3D Mark and running Firestrike Extreme for my benchmarks and all was fine. Never an issue since. As I recall Skydiver was written for mobiles or weak cards or something to that effect. I ended up with a second 980 FTW in January for SLI and still didn't have any issues other than the newer card had a much lower Asic # and my SLI benchmarks weren't as  strong as I'd hoped. More research taught me lower Asic numbered cards worked better when  water cooled and higher Asic numbered cards clocked better on air. Have you just tried running default settings running Firestrike Extreme? I'm not saying you haven't got a bum card but if it runs fine at stock clocks at default voltage, EVGA isn't going to do anything for you. They only promise that the card will run at spec out of the box. Any overclock you get is determined by the luck of the draw at the silicon lottery. I wish you well and hope you get this figured out but I personally have no gripes about my FTW's other than I can't get after market water blocks for them. Not that I need them water cooled as running games "optimized thru GForce experience, like COD AW, Battlefield 4 and Battlefield Hardline have never given me any temps on my lower, hotter running card more than mid seventies. Again, good luck my friend, but believe me my FTW's do nothing but kick butt and take names.

Running full fan speed doesn't help?
 
If you're brave, open up the card and re-seat the heatsink (after a new thermal paste job on the GPU of course). Not because of the GPU temps per-se, but because of the likelihood that POSSIBLY the cooler doesn't sit properly on the VRMs which (in my opinion) is the reason for the black screens. Otherwise, blacks screens at stock are of course reason for an RMA.
 
>>
Is there really nothing to do here except downclocking by 100Mhz and just deal with it or doing RMA and playing this GPU lottery again and again and again?
>>
 
Yes there might be sort-of a workaround which I used on my GTX 970. The first thing is, there DEFINITELY is a relationship between fan-speed/temps and black screens. This cannot be denied. Therefore run fan as high as you can tolerate.
 
The actual workaround is to mod your bios to use LESS voltage than you max. voltage available. (For example, my 970 does 1506 at 1.187V and doesn't require more. My max would be 1.212V. Since less voltage causes less power draw/less heat..it also reduces black screens.) This is what I observed here.Whether Afterburner with a GTX 980 would allow do DECREASE voltage a few notches without BIOS mod I don't know, but I would give it a shot.
 
>>
because i don't like my card running @70+°C all the time so i don't think its an heat issue.
>>
Well not a GPU heat issue for sure, but of course your VRMs might be overheating.
 
After a TON of research this is what I am currently believing in regards to that black screen problem.
 
 
>>
Now i played for about an hour+ at 1080p with vsync and no blackscreen
>>
 
Yes because Vsync usually means less load. But as soon as the cards get close to (as I observed) 190W-200W power draw they black screen. This is why less voltage can help since it will keep overall power draw lower. (And no, less V doesn't mean less overclock, if you're lucky you can get to 1500 with only 1.187V and no darned black screens). AND FOR STOCK speeds, even less voltage may be required!

Well it could be related to the VRM's due to not getting cooled properly. And all about the power it draws and heat it generates therefore could also be spot on. The thing is we shouldn't have to guess what the problem is in the first place because as far as i am concerned, it shouldn't happen at all. For the price we are paying we should be able to expect a product working as advertised and not have to troubleshoot, reseat heatsinks, lower power consumption or mess with the bios to get a stable card at all for that matter.
 
I know this can happen to any vendor, so i am not calling out EVGA specifically. I remember having to mess with the bios on my GTX680 Lightning from MSI to have it stable. But seeing how much reports there are about this issue specifically with either SC or FTW cards makes me think there is something wrong by design.
 
I already made my decision to run it downclocked until next week (so far its working) and do a cross RMA then. If the next card gives me the same problem i will RMA again and again until i have one that works.
 
Also, if todays rumors about the 980Ti releasing side by side with the Fiji next month at Computex are true, i will step up as soon as it is available. So here's hoping for that.
 
  

Might happen, who knows. Its not like i have a better choice tbh. Its either RMA at EVGA or RMA at the shop i got it from by now. At least i can cross RMA with EVGA while i would have no working pc if i would RMA through the shop. So long as i can step up to a 980ti sometime next month i don't mind anymore at this point. Hopefully it will not have any problems like this or a ridiculous low voltage block like this card.
 
 

Very informative post Samsander, thanks.
 
That seems indeed very weird. Why would they just not cool them at all? Shouldn't have been too hard to redesign the front plate to prevent this kind of issue, right? Or am i missing something here?
 
If it's a design flaw how could it have been overlooked in the first place? I doubt we will ever get an answer but I would like an official response anyway if possible. 
 
That's the second thing about this card that doesn't make any sense to me. First the hardware voltage lock way below reference and now this. I really regret getting this card.
 
Edit: Honestly, i think its worth opening a separate thread about Your findings but its up to You.
Edit2: It would be interesting to know if the SC front plate is also not cooling those VRM's at all. As far as i remember SC's are using the reference design for the board, right?

Sounds somewhat like my issue I had with my 980 FTW. I gave up after getting 3 bad cards and then decided to get a 980SC instead. It seems something is wrong with that card because out of all the 980's is the most that has the most threads about it in terms of issues.

wickedwayne
Step Up is now available for the 980ti, so i guess thats that. Hopefully it doesn't take too long. 

Doing the same thing as you, I just want to get rid of the 980 FTW. I have some of the same issues as you with my card. I had my system lockup multiple times on desktop. Swapped out my 980 with my old 760. Not a single issue since.