Consistent graphics driver crashing solely on ESO

Gazash · July 2014

Setting games settings back to D3D11, crashes the game immediately upon entering Tamriel. Setting the user settings to D3D9 loads into Tamriel only to crash within a few seconds of being in there...
So, still not working for me

Gazash · July 2014

Just thought I'd update you on what I've achieved...

I went into my UEFI to check overclock settings. And nothing was overclocked at all (which I thought strange).
So I went through the EZ Tuning Wizard. Set it up as per what I thought it should be. With the usage scenario set to gaming and cooling set to liquid cooling. It recommended an overclock of 14% on CPU and 48% on RAM. Selected the recommended settings and rebooted. Before trying anything else I ran 3D Mark, previously something unachievable, and these were my results:

All tests completed with no graphics driver crashes. Something that up until now I haven't been able to achieve.

Gazash · July 2014

And just to add to my previous comment, ESO now plays with DX11 as the default and I have had no problems at all since logging in after making these setting alterations to my UEFI BIOS.

Most confusing, but I will continue to monitor my system and play.

I appreciate everyone's help up to this point though

RinaldoGandolphi · July 2014

I would be very careful of using wizards and auto overclock methods. They tend to supply way to much voltage.

tons hardware had an article bout it, and found the CPU they tested bios auto overclockd on ended up suffering from electro-migration and the CPU degraded over a period of 6 months even with water cooling due to voltage which is a silent killer.

for example I have an FX 8320 oc to 4.4ghz at 1.36v the stock voltage on the chip is 1.380v so my overclock is stable but running undervolted. Your goal with oc is always getting the best clockspeeds with the lowest voltage possible, it takes more time, but your better off able to run it with less volts.

to be on the safe side I wouldn't exceed 1.30v on CPU voltage/vcore or 72 degrees Celsius on temps for long periods of time or you risk damage.

that's a great chip you got there, just be careful.

Saerydoth · July 2014

Gazash wrote: »

Ok so... I've still not solved this one, so here's what's been happening this week!

I’ve had my pc a week now and I’ve been experiencing on-going graphics issues since it arrived. When I’ve had time I’ve been trying to work around them and figure out what’s actually going wrong but so far I’m getting no where.

Spec is as follows:
Intel Core i7 4790K CPU OC to 4.4GHz
Corsair H100i Water Cooler
Sabertooth Z97 Mark 1 Motherboard
NVidia GeForce GTX780Ti 3Gb Video Card
16Gb Corsair Vengeance DDR3 2400Mhz Memory
Samsung 840 Evo SSD 250Gb
Seagate 7200rpm 2Gb HDD
24x DVD-RW
Corsair RM850W PSU
Running Windows 7 64bit

The problem I’m experiencing appears to be driver related. I’m mainly playing Elder Scrolls Online but the problem persists with any game I’ve tried to play so far. In Elder Scrolls, if I edit the User Settings file and set it to run on D3D9 (DirectX 9), I can get it to run, although it does have graphics glitches where it will drop output to the displays for a second or two, then resume. If I set it to run on D3D11 (DirectX11), as soon as I get past the character selection screen and load into the world, it crashes.
It then gives me an error in my Windows System Tray saying “Display Driver has stopped responding and has recovered. Kernel Mode Driver, 337.88 stopped responding and successfully recovered”.
I’ve been running the MSI Afterburner Software as a means of monitoring the graphics cards temperatures and usages, and it doesn’t appear to be out of tolerance anywhere which leads me to think it isn’t a hardware related issue.
Some recordings of recent gameplay:

If you look at the GPU Usage statistics, where you see the GPU usage drop is where it would drop display for a second or two and then pick it up again. But it’s quite frequent.
Even under the most strenuous graphically demanding situations I’ve encountered so far the GPU never went above 65°. I have generated a custom fan curve that have the fans come on with more power sooner, but it saved around 5° over the default fan control settings.

I have tried to run 3DMark 11, and the 3DMark Demo (Available from Steam).

3DMark 11 went as follows:
Installed and loaded the software. Set it to benchmark tests only, on centred view. Pressed “Start” and it immediately turned off display. Giving the monitor the “Input Not Supported” screen that floats about, so I’m guessing that’s a resolution issue? But I couldn’t correct it as I only downloaded the Basic version and didn’t upgrade to the paid version. So, with no results, this was an unsuccessful test.

3DMark Demo went:
Installed and loaded software. Clicked on “Run all tests”, and it started to work, then crashes out with the System Tray video driver error I had before.

So, the list of things I’ve tried to correct the problem.
1. Updating drivers using GeForce Experience.
2. Updating drivers manually by uninstalling drivers using Windows Task Manager first.
3. Uninstalling drivers using Windows, installing previous driver versions.
4. Installing latest Beta version driver.
5. Updating DirectX (But the installer wouldn’t let me because the version I have is equal to, or newer than the installer).
6. I’ve checked the XMP settings in BIOS. Already set to enabled so I left it at that.
7. Reinstalling my games.
8. Custom fan curve (though I imagine this is having no effect in relation to the problem I’m having).
9. I’ve checked DxDiag and it reports no problems.

So, to sum up. Won’t play games without crashing. Won’t run 3D Mark benchmark tests. Graphics driver crashes. And I’ve run out of ideas...

I should stress that this problem is not just limited to ESO as I'd previously thought. I tried to play Warframe today and the video card drivers kept crashing.

That "video driver stopped responding and has recovered" error is hardware. I have a nearly identical system to yours (only difference is 4770k instead of 4790k, different power supply, and different/more RAM).

I have the EVGA SC 780ti. My first one would do that ALL THE TIME, in various games (but was pronounced in Arkham Origins). Underclocking it by 30 MHz made it stable. It was swapped out under warranty and I've been completely fine since.

Elf_Boy · July 2014

Gazash wrote: »

@Elf_Boy‌ - I'm running 2x HP w19 monitors. Running at 1440x900, 60Hz on DVI cables for both. I'm running Windows 7 x64. There is nothing running in the background that would cause any notifications to pop up, other than the graphics driver failing. And the problem I was having occurred before I started using MSI Afterburner to monitor my system. I'm starting to think it may be a defective card also. I've already emailed the company that supplied the pc to me to get the ball rolling.

Thanks for the response guys, much appreciated!

If you have a spare cable around just as an experiment see what running with HDMI or a VGA cable does. Dont go buy one, not that much of a test. Also if it is not too much of a hassle try going to just one monitor and try them both singly.

This could eliminate the monitor and cable from the cause - unlikely but possible.

theyancey · July 2014

I am running the 340.43 beta driver and am having no problems. I did have a situation like this a few years ago. I sent my card back but that did not fix the problem. Turned out to be a defect on my MOBO.

Seravi · July 2014

My NVIDIA settings are as stated by CS. Have been since April. I have had upper level CS Tech guys trying to help with this and nothing works other than DX9. I've tested power, memtest, cpu test, latency tested and tested anything else that could possibly be an issue and nothing is bad, broken, slow, etc.

I've tried beta, and every driver from current back 10 levels, same results.

Elsonso · July 2014

I had this problem during the beta with an EVGA card. It was replaced under warranty with only a few weeks usage and that solved the problem.

I am thinking I will stay away from EVGA cards in the future, although they do have good customer support.

Abetow2531 · July 2014

I started getting the video issue with the last patch and only use the video settings in game. This is an ESO problem not a user problem. I have no trouble with any other game. You should look at your driver stack in your source code for ESO.

Gazash · July 2014

If you'll see my previous posts, you'll see that this problem isn't solely related to ESO but to other games also.

I know that using built in OC facilities can be a bit rubbish, but my computer is stable now. I've just played ESO for around 7 hours and have since had no problems with my graphics at all. (And that's with DX11 enabled in User Settings). I will look into the voltage ratings though, thanks for pointing that out.

I think my problem stemmed from from my system being badly synchronised. But since I've run the OC wizard in the UEFI, it has run perfectly so I can only assume that this has synchronised any instability issues I was experiencing.

RinaldoGandolphi · July 2014

No problem

you can use hwinfo to keep track if things, just alt tab out every so often and check your temps, look at your max temps and vcore, if you see it higher then 1.30-1.32 consider backing off voltage in the bios a little, could maybe use a 0.025 negative offset to keep it in the safe zone.

other then that, glad you got it stable

Lirkin · November 2015

This is happening to me now. Goes to black screen but sound and keyboard still work. I have to reload the game.

My card is not over clocked. My system is not overheating.