Graphix issue: Display driver xxxxx stopped responding and was recovered(TDR error)

AJack10600

Elite
Joined
Jun 13, 2005
Posts
2,732
Location
Switzerland
Society
SPU ¦ Smart Professionals United
Avatar Name
Larios Semai Niva
When playing EU I very often experience Screen freezes and Crashes due to this message:

Display driver xxxxx stopped responding and was recovered

This is a known error since Windows 7. Windows has this TDR feature that simply tries to protect your system from something. The problem is that this error can have like 1000 reasons for appearing...

the main ones being:

Poor Cooling
Problems with the power supply
Overclocking Issues
Bad System memory or incorrect memory timings
Defective PC Components

But also:

badly coded games...

So my question here is who else is having this kind of problem and what is your setup. Just trying to collect info on this in relation to EU.

Other games work fine on my PC which makes me wonder.
 
I had this problem also wuth vu10. Other games worked just EU crashed BSOD and the graphic reseting. I ran memtest+ and it was due to faulty memory. After i replaced it no errors what so ever. Proably EU pumps the full 4 gig full of data and causes some crashes in the memory.
 
You may want to post your own system specification and also temperature of CPU / system / graphics card when this is occurring. Also which graphic driver version you are using.

If you dont use a temperature monitoring program, Speedfan is good and free.


I've not had that error.
 
I had this daily for months and months - some days worse than others. And only happened while running EU.

Win7 64bit Professional Edition
ATI HD4870 graphics card - bought December 2009

No cooling issues, latest drivers, no defective components.

With that card I should be able to run in Very High.

But unless I have the 'Shading Quality' set to Medium or less, I'll get graphics error messages.

Now I'm fine - unless I go east of Argus. For some reason the pvp zone up there will give me graphics errors over and over.
 
1. Try using a Credit Card with 3d Secure.
2. Also, if you don't deposit often enough you may see "weird" errors like that. :wtg:

On a serious note, check your Event Logs... all of them, and try to see if it tells you which application caused it.

When did this issue began to appear?
 
I had that problem a while back... when I still had an under-rated power-supply and graphics card that overheated.
No problems since I upgraded psu and bought used 8800GT.
 
phenom 4 core cpu / ati HD4870 / 6GB ram ... all setings on high... grafic issues = most time no issues:yay:

To solve overheating, i use a strong extra fan!
 
I had this problem too before, with my old GTX260 Black Edition.

I just got the problem in EU of all the games i played, but i also got it in 3D Mark 06 not in vantage though(This was after VU10 btw).

Anyway, i did an RMA on the card and did recived a new GTX260 Black Edition....and got the same problem....I then did a RMA on that card too and asked the netstore i bought it from to give me an ordinary Gainward GTX260 instead. And i did get that, and after that no more problems....

Since then i have upgraded to a GTX470 and that is running without any problems.

The rest of the system are:

CPU: i7-860

Win Wista *64

4gb ram

PSU: Corsair 750w

Motherboard: Gigabyte P55-UD4
 
my win 7 laptop has:
c2d 2ghz
2gb ram
8600gs

i've had this problem only once or twice on my laptop.
 
i have exactly the same problem, pretty often....
It just happens in EU, no other game.
maybe we can figure out which hardware component(s) causes this problem
...
System:
CPU: Intel I7 920 (overclocked to 3,2GHz)
GPU: ATI 5870
Motherboard: Asus P6T Deluxe V2
RAM: 6GB Corsair Dominator GT
PSU: Enermax 640W
liquid cooled (never more then 40°C)

all games and benchmarks are stable, so im pretty sure some hardware components are not supported very well.
 
I did a lot of checks...

I doubt it has to do with my system. I tried Overclocked and on standard settings. I also have 5 Cooling fans with Fanspeed controls and Temperature sensors in my case. I can run at 30C easy.

I have run Memtest and Prime95 with no issues at all. Issue log just confirms that the Graphix Driver has caused an error. amdkmdap is the driver file.

I was playing with Fan Speed on the Graphix card yesterday evening, bumping it up to max and was not getting any crashes as long as the GPU did not go over 60C. But For a GPU, 60C is not much. It can go up to 80 easy... So something is really odd here.

Mega ( I tried ATI CCC and GPU-Z to check temps )

I have all the latest AMD drivers installed.

AMD Phenom II X4 965 BE C3, 3.4GHz;
Zalman CNPS10X;
MSI 870A-G54;
Corsair 2x4GB, DDR3-1333;
Sapphire HD 5870 Vapor-X 870MHz;
Western Digital Caviar Black, 64MB, 1TB,
Termaltake 650W PSU

Windows 7 64 Bit

What I have not tied yet is taking the Clock speed of the Grapix Card down. This is an OCed Vapor X. The Standard Clock speed for the HD5870 is 850 MHz not 870. Maybe I can try taking that down... although this is silly if I spend money for a better card...

As I said, this is a EU problem for me. No other game produces this...
 
Last edited:
I did a lot of checks...
....
....
As I said, this is a EU problem for me. No other game produces this...

I heard from a lot ppl, that they only have this problems while playing EU.
 
Long shot, but have you run the "repair tool" to check for any corruption in EU files?
If one texture is corrupt in just the right way, it can turn into an exploit for a buffer overflow or something of the type.
 
Long shot, but have you run the "repair tool" to check for any corruption in EU files?
If one texture is corrupt in just the right way, it can turn into an exploit for a buffer overflow or something of the type.

Nope, might give this a shot...
 
It was after a upgrade of my graphics card that I got this problem.

Turned out it was my PSU not generating enough power for new card. Remember that a PSU loses poweroutput after time.
Calculate your power req. here and don't forget to include the capacitor aging % at the bottom. I use about 10% decay for every year(so a 2yr old PSU loses 19% capacity).

Ofcourse this is just 1 possible source, could be plenty others as you mentioned. :laugh: gl
 
I get this everytime I turn off my comp.. then when I turn it on again this msg will come in any game I start playing.. its like the drivers are fucking up.. but I have the latest drivers from nvidia.. overheating in my tower is not even possible.. and the drivers should be ok..

EU dont crash.. but the screen just go black for a few seconds.. then it is back to normal.. same in all other games..
 
I had this daily for months and months - some days worse than others. And only happened while running EU.

Win7 64bit Professional Edition
ATI HD4870 graphics card - bought December 2009

No cooling issues, latest drivers, no defective components.

With that card I should be able to run in Very High.

But unless I have the 'Shading Quality' set to Medium or less, I'll get graphics error messages.

Now I'm fine - unless I go east of Argus. For some reason the pvp zone up there will give me graphics errors over and over.

you need 4 gig of ram to run very high
 
It was after a upgrade of my graphics card that I got this problem.

Turned out it was my PSU not generating enough power for new card. Remember that a PSU loses poweroutput after time.
Calculate your power req. here and don't forget to include the capacitor aging % at the bottom. I use about 10% decay for every year(so a 2yr old PSU loses 19% capacity).

Ofcourse this is just 1 possible source, could be plenty others as you mentioned. :laugh: gl

ROFL... as per this website I would need a 1000 Watt unit for my PC... pfffffff can't be serious...
 
ROFL... as per this website I would need a 1000 Watt unit for my PC... pfffffff can't be serious...

Well, I'm guessing you did something wrong. Did you make sure you selected single socket at the top? Your CPU has 4 cores but sits on 1 socket.

Because when I enter your specs with 2 HDD and 2 DVD drives plus 2 additional PCI cards, 5 fans plus leds and a 1 yr old PSU I come to 699W.
 
you need 4 gig of ram to run very high

I have that, so not the issue.

If you google around, you'll find that this driver display error has been an issue with Win7 for 3 or 4 years now.

I've tried every sensible solution I can find suggested on various websites. None of them make a blind bit of difference.
 
I have that, so not the issue.

If you google around, you'll find that this driver display error has been an issue with Win7 for 3 or 4 years now.

I've tried every sensible solution I can find suggested on various websites. None of them make a blind bit of difference.

i got a 4890 and can run very high with no problams windows 7 same as you
 
Constant graphics driver crashes are usually caused by 1 of 3 things. Thermal problems, psu problems, or overclocking problems(anything, mb, ram, cpu, gpu, etc).

Ive run many graphics cards in many systems over the last dozen years, as well as helped out a lot of other people. And random constant driver graphics crashes are almost always one of those 3 things. (ive seen weird stuff before as well, for instance an agp graphics card that had really bad artifacts under default bios settings, but manually configuring the voltage lower then default and cranking up the amperage rating on the agp slot and the card was rock solid with no artifacts, the card wasn't the problem in this instance it was the motherboard not providing stable power to the agp slot)

Of course overclocking is also a biggie. A lot of people think they have a stable overclock when they don't. A truly stable cpu overclock needs to have the processor pegged for days on end without crashes. A truly stable gpu overclock needs the same. Different applications will stress the hardware in different ways. Just because its stable in one app isn't enough. Prime stable isn't enough for a cpu, looping 1 graphics benchmark isnt enough for a gpu. Even days on end isn't enough, but usually at that point you can rely on it.

If you had to do a voltage change to get to your overclock that's a definite red flag. Voltage increases are the fastest way to kill hardware. And are usually not worth it; especially if you plan to keep your hardware for more then 1-2 years. Just because an overclock with a voltage mod was stable 6 months ago doesn't mean its stable today. Ive killed hardware before with voltage increases even small ones. For instance a motherboard was stable for a year with a small voltage increase(0.1v) and a large bus clock increase(30%), then it would refuse to post unless i reset the bios on each boot. After that i could increase the clock rate and it was rock solid till i turned it off, a few months later even that didn't work and it would refuse to post completely.

As an example of things that might be harder to trace... Ive seen ram modules with 1 bad bit cause symptoms that made me think the graphics card was at fault before. Normal system operation had no problems. Nothing else seemed to crash except 1 specific game i was playing. And it was random. Which made me think it was a thermal issue. Eventually i found out a single page on 1 ram module was causing it. Using a utility to lock out the bad page of memory just after booting and bingo the system was 100% stable again. I continued to use that bad ram module for 3 more years without problems as long as the bad page was allocated and not used.

The PSU specs stated in this thread seem decent enough. Tho that's that doesn't mean they aren't failing and its not easy to test that unless you have some electronics expertise. Id give advice on what to look for but since some of the components are potentially lethal if touched when charged, im not going there.

For thermal problems, don't trust the temp readings in a driver or a program etc. Crack your case and look at the thing. Is the fan spinning, is the heat sink caked with dust. If the fan isn't spinning, you've probably nailed your problem. If the heat sink is caked with dust time to clean it.

The first thing id tell you to do with a constant driver crash problem is reinstall the driver.Next is check the fan and heat sink. Next is stock clock on everything if you are overclocking, next would be to underclock your graphics card by about 20% and max the fan speed if you can control that. After that woudl be to try an older driver if its a recent problem.

Ideally you would want to swap out pieces and see if you still crash. But if you only have 1 system that doesn't work.

Unfortunately there are a lot of things it can be, and without spare hardware to swap it can be hard to diagnose sometimes.
 
Last edited:
Constant graphics driver crashes are usually caused by 1 of 3 things. Thermal problems, psu problems, or overclocking problems(anything, mb, ram, cpu, gpu, etc).
...

This post requiers in my view a Sticky because it is a excelent summary for every Gamer with Graphik problems!!

Cant rep you need to spread first
 
@ Atraie,

Raffaele suggested testing the Ram modules individually which is what I will do. Alltogether in Memtest they ran fine with no problems.

The next thing I will do is reduce Clockspeed on my Graphix.

I really am singling out any other issue. It's either Ram or Graphix related.

My temps are real, I have Temp sensors all over my casing on top of the application ones.

I'll post here again after my checks.
 
Back
Top