GTX780 SLI Display driver crashing [HELP]

sub-80

Senior member
Jan 11, 2014
259
4
81
Hey guys;

The display driver is crashing ("Stopped working" in a white notification box below) or the game goes black and jams, but nothing is wrong with the os.

This is my PC specs in the pic below: (SLI enabled "Maximize 3D" set in Nvidia Control Panel.)
0ienV2R.jpg


  • VGA: Gigabyte GTX780 Windforce GHz edition. [Base/Boost:1019/1071MHz and memory clock:6008MHz] (Strange thing is that it was set out of the box when I received it to 1020/1072MHz and memory:6008MHz so I kept it like that.)
  • PSU: Seasonic X-series 1250W (Model:SS-1250XM Active PFC F3)
Have 2 monitors"

  1. LG Flatron W2363D Monitor[23", 120Hz, 1080p, TN panel, 3D, 3ms (GTG)]
  2. LG 32LH70YR LCD TV [32", 100Hz, 1080p, LCD panel, No 3D, 2.4ms (GTG)]

  • I use LG Flatron monitor for gaming and LG TV for movies.
  • Ran the game in the optimal setting suggested by GeForce Experience (if the native resolution set by Geforce Experience is higher than 1080p, it is lowered to 1080p)
  • Ran the GPU-Z and HWMonitor but the CPU remained below 50C and GPU ran in the 60sC
Ran 3DMARK Fire Strike Extreme v1.1 Below are the Results:
From app:

o1vktrC.jpg


(Graphic driver not approved)


From website:
yzKd5HP.jpg





Reinstalled the Nividia driver but no affect. So what should I do?


Thanks;

UPDATE1: I did the custom and clean install then I went and played call of duty for 3 missions and ran 3DMARK but noticed that I didn't have SLI enabled. So back again to testing.

UPDATE2: Played call of duty at the same setting but it crashed below are pics of notification and warning:
A2wwTiR.png


g4v2svi.png


UPDATE3: Tried running with an older driver choose at random (344.65) Was alot worse and when I went back to (347.09) was the same as (344.65) so I deleted the driver via control panel and had windows install the driver.

Played COD :AW with SLI disabled played along time before the game jammed but there was no notification of the Driver crashing.

Going to to try remove one VGA and play to see if it crashes then try the other alone.

UPDATE4: Tried the DDU but had no affect. so I went with the following tests:

OCCT 4.4.1: CPU:OCCT
[Duration:1hr/Test Version:64bits/Test Mode:Large Data Set/Number of threads:8/Error Check]
Result: Passed

Furmark V 1.12.0.0: Burn-in Benchmark 1080p 15min MSAA 0X
%: 40 (Manually Stopped) SLI Enabled
###################Temp Fan
P (Primary) Slot: GPUA ####90C 99%
S (Secondary) Slot: GPUB##80C 65% Score:N/A

Furmark V 1.12.0.0: Burn-in Benchmark 1080p 15min MSAA 0X
%: 40 (Manually Stopped) SLI Disabled
###################Temp Fan
P (Primary) Slot: GPUA ####80C 66%
S (Secondary) Slot: GPUB ##30C 17% Score:5857

Unigine Valley Benchmark
Extreme HD [API: DX11, Quality:Ultra, Stereo3D Disabled, Monitors:Single, MSAA:x8, Fulllscreen, 1080p] SLI Disabled
#######Temp Fan
PSlot:GPUA 70C 52%
SSlot:GPUB 30C 17%
FPS:61 Score:2570

Unigine Valley Benchmark
Extreme HD [API: DX11, Quality:Ultra, Stereo3D Disabled, Monitors:Single, MSAA:x8, Fulllscreen, 1080p] SLI Enabled
########Temp Fan
PSlot:GPUA 72C 57%
SSlot:GPUB 65C 47%
FPS:84.5 Score:3536

Unigine Heaven Benchmark
Custom [API: DX11, Quality:Ultra, Tessellation: Extreme, Stereo3D Disabled, Monitors:Single, MSAA:x8, Fulllscreen, 1080p] SLI Disabled
########Temp Fan
PSlot:GPUA 70C 52%
SSlot:GPUB 30C 17%
FPS:54.4 Score:1375

Unigine Heaven Benchmark
Custom [API: DX11, Quality:Ultra, Tessellation: Extreme, Stereo3D Disabled, Monitors:Single, MSAA:x8, Fulllscreen, 1080p] SLI Enabled
########Temp Fan
PSlot:GPUA 78C 63%
SSlot:GPUB 68C 57%
FPS:101.2 Score:2548

Furmark V 1.15.1.0: GPU Stress
[Fullscreen/1080p/MSAAx8] Run 10min
#######Temp Fan
PSlot:GPUA 81C 67%
SSlot:GPUB 76C 61%

COD :AW SLI Enabled Temp:55C&51C V:0.94&0.94 Crashed after ~15min

Removed GPUB & kept GPUA in PSlot.
Screen gave Overclock failed when starting pc so I set it to Default values (Set CPU and memory to stock) PIC below:
v704M2b.png



Unigine Valley Benchmark
Extreme HD [API: DX11, Quality:Ultra, Stereo3D Disabled, Monitors:Single, MSAA:x8, Fulllscreen, 1080p] SLI Disabled
#######Temp Fan
PSlot:GPUA 66C 48%
SSlot:none none none
FPS:59.4 Score:2486

Unigine Heaven Benchmark
Custom [API: DX11, Quality:Ultra, Tessellation: Extreme, Stereo3D Disabled, Monitors:Single, MSAA:x8, Fulllscreen, 1080p] SLI Disabled
#######Temp Fan
PSlot:GPUA 66C 48%
SSlot:none none none
FPS:55.9 Score:1408

3DMARK FireStrike Extreme Score: 4603

Furmark V 1.15.1.0: GPU Stress
[Fullscreen/1080p/MSAAx8] Run for 10min
########Temp## Fan
PSlot:GPUA #70C###54%
SSlot:none #none##none

GPUA removed and replaced with GPUB

Unigine Valley Benchmark
Extreme HD [API: DX11, Quality:Ultra, Stereo3D Disabled, Monitors:Single, MSAA:x8, Fulllscreen, 1080p] SLI Disabled
#######Temp Fan
PSlot:GPUB 69C 51%
SSlot:none none none
FPS:61.2 Score:2559

Unigine Heaven Benchmark
Custom [API: DX11, Quality:Ultra, Tessellation: Extreme, Stereo3D Disabled, Monitors:Single, MSAA:x8, Fulllscreen, 1080p] SLI Disabled
#######Temp Fan
PSlot:GPUA 69C 51%
SSlot:none none none
FPS:55.9 Score:1408

3DMARK FireStrike Extreme Score: 4684

Furmark V 1.15.1.0: GPU Stress
[Fullscreen/1080p/MSAAx8] Run for 10min
#######Temp Fan
PSlot:GPUA 71C 56%
SSlot:none none none

COD :AW Played for an hour and no crashes

PSlot:GPUB & SSlot:GPUA

3DMARK FireStrike Extreme Score: 7801 & 7818(Second time)

COD :AW Crashed after a few minutes


Went into the BIOS and into the Hardware Monitor
CPUV: 1.240V3.3V: 3.344V
5V: 5.088V
12V: 12.152V

For 10min no voltage fluctuation

Samsung SSD magician (SSD Management Software)
Set Max Performance which set virtual memory to maximum value

UPDATE5: Disabled sli and removed sli bridge and I played 2hrs twice with no crashes or glitches. I bought a used ASUS ROG sli bridge (assuming) and it did the same thing when I connected it.
 
Last edited:

Keysplayr

Elite Member
Jan 16, 2003
21,219
56
91
Did you choose "Custom Install" and then "Clean Install" when installing the Nvidia driver?
If not, give that a try.
Also, list the complete specs of your PC including make and model of your power supply and your mobo.
 

AdamK47

Lifer
Oct 9, 1999
15,846
3,638
136
It can often times be caused by CPU or memory instability. It's not necessarily triggered by the graphics sub system.
 

sub-80

Senior member
Jan 11, 2014
259
4
81
Did you choose "Custom Install" and then "Clean Install" when installing the Nvidia driver?
If not, give that a try.
Also, list the complete specs of your PC including make and model of your power supply and your mobo.

The motherboard specs and ram are available in the first pic (on the left side). MY PSU is Seasonic X-1250W Gold 80 Plus (Model:SS-X-1250XM PFC F3)

It can often times be caused by CPU or memory instability. It's not necessarily triggered by the graphics sub system.
I have 16GB of 1333MHz Ram and the SSDs are managed by samsung ssd management system which is set to max performance.
 

Hauk

Platinum Member
Nov 22, 2001
2,806
0
0
GPU overclocks, or Boost ramping max GPU clock too high can cause driver crashes. I had 780Ti SLI where temps, psu, drivers were all good, turned out it was my overclock. Overclocking?
 

sub-80

Senior member
Jan 11, 2014
259
4
81
it is a Gigabyte GTX780 Windforce GHz edition. Baseclock/Boostclock:1019/1071MHz and memory clock:6008MHz. Strange thing is that it was swt out of the box when I received it 1020/1072MHz and memory:6008MHz so I kept it like that.

So I haven't overclocked the cards they are factory overclocked.
 

Hauk

Platinum Member
Nov 22, 2001
2,806
0
0
Boost could ramp GPU higher given other favorable variables; curious if setting nv CP power management mode to max performance helps. GPU should cap at 1020, curious if stability improves. Nice PSU, same here, other stuff looks good.
 

digitaldurandal

Golden Member
Dec 3, 2009
1,828
0
76
Set everything you have to stock if you have not already.

Attempt again and see if the issue persists.

Remove one card and attempt again, and again with the other card.

This will narrow down the issue.

If it occurs with one specific card, it may be that card. If it occurs with both cards it is something else, if it occurs only when two cards are present, it may be power related.
 

Puffnstuff

Lifer
Mar 9, 2005
16,256
4,930
136
The last time I experienced a problem like that was when my sli bridge went bad. You might try removing it, cleaning the contacts on the cards, and then reseating it.
 

Deders

Platinum Member
Oct 14, 2012
2,401
1
91
Does this happen with the previous drivers after a clean install? It may just be this version of the drivers.
 

sub-80

Senior member
Jan 11, 2014
259
4
81
Tried running with an older driver choose at random (344.65) Was alot worse and when I went back to (347.09) was the same as (344.65) so I deleted the driver via control panel and had windows install the driver.

Played COD :AW with SLI disabled played along time before the game jammed but there was no notification of the Driver crashing.

Going to to try remove one VGA and play to see if it crashes then try the other alone.
 

PhIlLy ChEeSe

Senior member
Apr 1, 2013
962
0
0
How are you removing the old drivers? Deders asked if you did a clean install? Your not removing the old drivers completely.............
 

sub-80

Senior member
Jan 11, 2014
259
4
81
Steps
Chose custom (Advanced) then checked clean install and installed.

This is what I did when I installed 344.65 and when I went back to 347.09
 

Hauk

Platinum Member
Nov 22, 2001
2,806
0
0
Any luck OP? Have you tried benching one card at a time?
 

AdamK47

Lifer
Oct 9, 1999
15,846
3,638
136
I know the immediate knee jerk reaction is to blame drivers or the card. I still stand by the CPU / memory sub system instability potentially being the culprit. I've encountered NVidia display driver crash/recovery problems many many times in my overclocking adventures while using Vista/Win7/Win8-8.1.
 

Rvenger

Elite Member <br> Super Moderator <br> Video Cards
Apr 6, 2004
6,283
5
81
Someone here recommend to use DDU to uninstall the drivers? Uninstalling from control panel does not uninstall drivers properly. Use DDU and the program will remove the drivers in safe mode.
 

RussianSensation

Elite Member
Sep 5, 2003
19,458
765
126
I would do stability testing on the CPU and RAM and then each GPU separately. If these all pass, load the CPU and GPU simultaneously and then monitor 12V rail voltages on the PSU.
 

Hauk

Platinum Member
Nov 22, 2001
2,806
0
0
Dude you don't provide enough info, but curious I'd you've benched each card seperately, or tried capping GPU clock using nv cp, both easy enough,
 

sub-80

Senior member
Jan 11, 2014
259
4
81
The info is at the 1st post, which I update.
I benched each card alone and both at different slots (primary & secondary pcie slot) all are at stock values
 

Hauk

Platinum Member
Nov 22, 2001
2,806
0
0
My bad, so GPUa oc fail on stand alone bench, fail again w/ GPUa installed in secondary slot..