Hi guys,
I'm wondering what exactly I should do to solve my problem?
 
Short history: my 2500k or my p-67 pro decided it wanted to retire early in March. One day Rainbow6 Seige started throwing black screens, followed by the message, "no input device" from my monitor. At first it happened once in a while, then frequently. It got to the point where I couldn't even play a round. Also, the system would no longer hold an overclock, so I updated with a new psu, new mobo, new cpu, new ram, new ssd: Seasonic Platinum 760, Asus z170 pro gaming, 6600k, g.skill 3200 cl14, Samsung 850 pro.
 
Everything was fine, for ~ a month. Since Rainbow6 went to crap, I started playing SW Battlefront. At first I thought my problems were solved. I would play a round or two here and there while testing my overclock stability. However, I started seeing a lot of directx crashes, tdr's, and just about 20 minutes ago, the dreaded no input. At first I thought it was my overclock, but it's now happening running the cpu at stock. I've never overclocked the card (didn't need to), so there's nothing to undo on that end.
 
The system is stable in all other respects. Passed 8 hours memtest with zero errors at 3200MHz cl14 at stock and at 4.5GHz. The system has also passed RealBench, OCCT, and Prime95 v26.6 stock, at both 4.5GHz and 4.6GHz for an hour each.
 
The only other games I have installed are Wolfenstein Old Blood and Tribes Ascend (just moved to Seattle, all my stuff is in storage). I can play these for hours stock or overclocked. I've tried every single driver set (beta included) from 358.50 to the current 365.10's. Battlefront crashes with them all the same way: sometimes a black screen, sometimes the game crashes with no message that I can see, sometimes directx error telling me I don't have a card... which... yeah...
 
Other tools which have zero issues are Valley, Cinebench, Aida64, and 3DMark. I noticed that there may be an issue with gtx 780's and Battlefront, but can't be sure because internet: https://forums.battlefront.starwars.ea.com/discussion/11617/regarding-dx-crashes-for-nvidia-780-780ti-users
http://forums.battlefront.starwars.ea.com/discussion/8754/repetitive-direct-x-crash-nvidia
However, these are my exact issues.
 
Temps are not an issue. Recently installed a Noctua NH-D15 with both fans; stressing the cpu with OCCT or Prime max temps are 70C no matter the voltage I've applied. I also run Precision every once and a while and gpu temps are always ~ 60C.
 
What test can I run to make sure my card isn't bad that I haven't tried? Has anyone deal with this Battlefront issue?
 
Thanks in advance for your time.
 
 
 

Hi Sajin,
 
Thanks for the reply.
 
Just ran a quick game of Battlefront in debug mode with the cpu at stock with 365.10's. It crashed after ~ 10 min with the DirectX function error. What now?
 
*edit*
 
One of the suggested tips is to reinstall the c++ redist, which I seem to do almost daily now. Also, it's been my experience that once this crashing starts, the only thing that will make it stop is a driver reinstall, and then, that only works for a little while... and by that I mean playing more than 1 round at a time, which is bad because then I never get the play the other side of the map.
 
**edit2x**
 
Since my stuff is in storage, do you have any idea where I can get the driver that came with the card? The earliest I can find on nvidia's site are the 352.86's.
 
***edit3x***
/sigh....
 
I suppose I should also mention that the last few times I've uninstalled drivers the monitor loses signal right after I'm asked to reboot. Just happened again now. Have to press the reset button to get it to reboot... which is met with a "windows didn't shut down properly" message.
 
I'm getting the impression this card is not long for this world.

When overclocking I was running HWMonitor. I tried with it off after confirming temps were safe, same issue. Oddly, the one thing that has allowed me to play the longest without crashes is running Precision... it's almost like it knows I'm keeping an eye on it so it behaves. Although that doesn't last long either. On or off same thing. Just reinstalled the 356.10's without experience; I'll give those another whirl after dinner. I'll also try the new 365.19's and the 88's you linked (thank you!).
 
**edit**
 
365.10 no experience: crash after 10 min.
347.88 no experience: driver 358.91 or later required to play game message.
365.19 no experience: internet crapped out ~ 5 min in. i'm in wallingford seattle and a raccoon got into sub station the other night and killed power for 40k. ever since we've been having strange electrical issues, losing the internet and having the house alarm go off happens about every 3-5 hours.
 
when it rains it pours.
 
**edit2x**
 
I suppose I should mention that when uninstalling the 365.10's that I lost signal again... this time before I was asked to reboot... the screen just went black on me and the system light turned orange. That's got to indicate something very wrong, right? Stuff like this just never happened with this card before.

So you had no crashes at stock cpu/gpu clocks while running 365.19?

Hmm. Maybe the issue was finally fixed in driver 365.19?

Fumigator

Sajin
Hmm. Maybe the issue was finally fixed in driver 365.19?

I think I misunderstood? This was all done in debug mode, so "stock clocks" = reference. Haven't mustered the courage to try regular clocks...

Oh, ok. Try regular clocks on the gpu and stock on the cpu next.

Fumigator
What's the verdict?

Time to rma the card or buy a new one.

Well, it turns out the gpu was bad after all. Did the rma (cross shipped) and I've been putting the new card through quite the workout all week. Not a single issue, game crash, or error message at any clock speed. Kinda sad... I loved that 780. On the bright side, evga sent me a 970 sc, so I think I'll be able to hold off on a 1080 until the Ti's come out.
 
So, a big thank you to evga. This is expressly the reason why I keep buying evga stuff. Now hurry up with the 1080Ti's. :-)

 
Glad you got fixed up.