system:
Core i7 6850K @ 4.0ghz
64GB Corsair Dominator Ram 3000mhz Quad Channel
Asus Rampage V Edition 10
Win10 Pro
Samsung 950 PRO 512G M.2 PCie (OS drive)
Samsung 850 EVO 1TB w/ Rapid Enabled (Data Drive)
2x Evga GTX 1080 SLI (Old school dual connector bridge that came with the board while waiting on 80mm HB to arrive) (Latest drivers)
Corsair AX1200i PSU
 
3440x1440 @ 100hz resolution (Predator x34) (DisplayPort)
 
Heres the issue. I have tested both cards in SLI, both cards individually and the problem occurs on both cards. I tested heaven uningine worked fine 6 hrs, I can play The division maxed, I can play Mad Max maxed. But in Guild Wars 2 within a few minutes I get black screen that is only able to be fixed by hitting reset key. Happened in Elite Dangerous too, but I managed to fix it there by limiting the framerate to 100 via vsync (unlimited is about 285 fps average in Elite Dangerous).
 
I have done full CPU stress tests, full Memory tests. Changed every power cable twice going to my AX2000i PSU as well as the modular ports power the cards and I cannot seem to get a handle on this. Why is games like Elite Dangerous and Guild Wars 2 unplayable on this system but I can manage The Division and much more demanding games just fine.
 
I have tried extremely aggressive fan curves as well to no avail.
 
Could I really have 2 bad cards? This system is 6 days old and has fresh install of fully updated win10 x64 pro, and all latest drivers for all devices.
 
Thanks in advance for any help.

Case fans do yes.

m0nkjee
Case fans do yes.

I think Sajin refers to GPU fans, no case fans.
 
The additional clocks comes from GPU boost, so I think is normal.

 
I'm having no problems with GW2 and my single 1080 FTW Hybrid, although GW2 is more CPU intensive than many games rather than GPU intensive. Boost clock at 2025 is very good, most would consider that lucky. My card boosts to 1987 on default settings.
 
I'm wondering if SLI is the issue? Try going into the NVidia Control Panel > Mange 3D Settings > Program Settings, and turn off SLI for GW2 and put the clocks back up to default and see how it goes.
 
* edit: my hybrid started the black screen 100% fans issue soon after. :(
 

Nereus
 
I'm having no problems with GW2 and my single 1080 FTW Hybrid, although GW2 is more CPU intensive than many games rather than GPU intensive. Boost clock at 2025 is very good, most would consider that lucky. My card boosts to 1987 on default settings.
 
I'm wondering if SLI is the issue? Try going into the NVidia Control Panel > Mange 3D Settings > Program Settings, and turn off SLI for GW2 and put the clocks back up to default and see how it goes.
 

1987 and stable? that's fantastic, but its still out of range of what its supposed to be. According to http://www.evga.com/Products/Product.aspx?pn=08G-P4-6286-KR The product page, at the bottom specifications say that the boost should max out at 1860. So If black screening occurs at 2025 and now that I have downclocked it to 1860 and it does not occur it's obviously because the cards shipped with incorrect clock speeds. That being said, I didn't slowly downclock 10-20mhz at a time, I just dropped it down to its intended which for me was a reduction of 155mhz. I can also say that performance loss is unnoticeable, all games seem to get same fps as when it was boosting to 2025 

If I do RMA, how long will I be without graphics cards (From Canada) on average?

In my opinion, downclocking the cards has actually revealed an issue. All of the 1080 largely boost to the same frequencies lets just get that out there. Your FTW was boosting to where my classy boosts and the same as others with SC and FTW DT cards for that matter. I would setup an RMA for both cards explaining to EVGA exactly what you told us here. I dont think the card should've made it through QA/QC if the boost wasn't stable.

Same thing here - Got my system last week friday, build it up on saturday, spend 3 days of awesome gaming then suddenly yesterday my card crashed 3 times: black screen and GPU fans on high revs. Only thing that i have done was installed Presision X OC to monitor temps and made the fan curve abit more aggressive (card would stay cooler). And also my card seems to be boosting to the same 2025Mhz readings. I've contacted EVGA and they told me to RMA the card. After this I contacted my local supplier and they said they will excanse my card to a new one on the next incoming batch on FTW cards. I would have used the cross-ship or "advanced rma" option (what they called it in europe) but the using a currier to sent my current card to them from Finland in point of a price is... well to put in one word: maddness =(  Well currently Im just hanging here and praying the next batch of FTW cards will arrive to the local supplier.

StealthArsenal
In my opinion, downclocking the cards has actually revealed an issue. All of the 1080 largely boost to the same frequencies lets just get that out there. Your FTW was boosting to where my classy boosts and the same as others with SC and FTW DT cards for that matter. I would setup an RMA for both cards explaining to EVGA exactly what you told us here. I dont think the card should've made it through QA/QC if the boost wasn't stable.

+1

I'm thinking that if you never moved the +gpu and +memory speed sliders above 0, and you had to underclock the cards in order to keep them stable, and by that I mean dial in -speed and -memory, then you have defective cards. The fact that it was running above the advertised clock speed is perfectly normal - that's how Boost 3.0 works. The defect could be something simple like the cooler not mounted very well to the gpu or something, who knows. Have you tried removing the cooler and remounting with new thermal paste?
 
Otherwise, if you got the black screen accompanied by sudden 100% fan speed then you definitely have a defective card, which EVGA already acknowledged, as pointed out by an earlier poster.
 
Bottom line though, you should not have to deliberately underclock your cards.

sethleigh
I'm thinking that if you never moved the +gpu and +memory speed sliders above 0, and you had to underclock the cards in order to keep them stable, and by that I mean dial in -speed and -memory, then you have defective cards. The fact that it was running above the advertised clock speed is perfectly normal - that's how Boost 3.0 works. The defect could be something simple like the cooler not mounted very well to the gpu or something, who knows. Have you tried removing the cooler and remounting with new thermal paste?
 
Otherwise, if you got the black screen accompanied by sudden 100% fan speed then you definitely have a defective card, which EVGA already acknowledged, as pointed out by an earlier poster.
 
Bottom line though, you should not have to deliberately underclock your cards.

This.
 
If you had to underclock the cards (entering a negative value) then the cards are defective and you will need to RMA them.

m0nkjee
think you guys missed my latest post last night, and how the cards are now magically working....
 
as for PSU, Corsair AX1200i, has passed all corsair link testing. It is connected to the motherboard via the primary ATX power cable and an additional 6 pin connector above cpu.

It doesn't matter if it's working or not. You experienced the issue and EVGA has confirmed it's a hardware problem. This means the issue may occur 2 years or 2 months down the line.

sethleigh
It's also possible you've got a weak power supply, or you didn't plug in all of the power cables from the PS to the motherboard (the ones that supply extra juice to the PCI slots), or something like that.
 
One thing you could try is putting all the sliders like +gpu speed and +memory speed back to 0 and moving the power limit slider to 120%.
 

+1

A weak or poor quality PSU can definitely cause problems with a card drawing this much power. That's why I always stick with SeaSonic or the upper end Corsairs. Even if there is an underlying issue with the FTW cards, a weak PSU could cause the problem to rear its ugly head more frequently whereas if you had a great power supply - you might never see any issue.

My EVGA 1080 FTW Hybrid (same card as the normal FTW, just has the all-in-one watercooler attached) had the same problem, would black screen with fans at 100% once in a while while playing World of Warcraft.  Was getting around 37c under load at 2025mhz default boost.  Like others have said it's not heat related.  (it might be in-game physics related)
 
I contacted EVGA & they gave me a free advanced RMA, the issue is hardware related.  They were very friendly & they were able to ship another card to me while I continued to use my defective one so I did not even go a single day without a card.  Had to game at like 1700mhz for over a week while waiting using the debug method below, but it didn't blackscreen anymore!  The new card came with a shipping label for me to ship my old one back so it did not cost me anything either.  Got the new card several days ago, was quite a bit warmer at around 44-45c under load at 1925-1988mhz default boost.  This one like the last is 100% quiet both when under full load & when idle due to water cooled.  Have not had the problem anymore so far, 100% stable!  I plan to buy their advanced RMA on this replacement very soon, was very happy the way it worked in the end, especially since I was not without a videocard while waiting!  Like the other thread says, they did identify & fix the issue on new cards they replace.  This issue only affects a small number out of all the ftw hybrids & ftw's out there.
 
Also while waiting, enabling the Debug mode completely fixed the black screen 100% fans issue while I waited for the new card.  Clock's will not boost at all, but was 100% stable which worked for me.  To enable Debug Mode go to Nvidia Control Panel, click help, then check "debug mode".  Very rarely when rebooting it would "uncheck" itself so always make sure it's checked when gaming if you have this issue.  The only way I think to uncheck it once you get your new card is to reinstall the drivers since the option becomes forced/greyed once activated.  Hope this helped someone.

I would just like to say, EVGA truly is the king when it comes to support and honoring their products. I doubt I would have received the kind of help and fantastic treatment from any other manufacturer.
 
They gave me free advanced RMA for both cards and get this, I am in Canada where Advanced RMA is not even generally offered by EVGA. That's how top notch this service was!