I'm getting VERY sick of this...

G

Guitarrasdeamor

2[H]4U
Joined
Apr 5, 2008
Messages
3,622
Alright, whenever I'm playing L4D, and this seems to be the usual game it happens in, but it has happened in GTA IV as well, my display driver stops responding. The frame stays on screen for a while, then goes off, then throws me back into the game with the little taskbar notification that it has recovered. Well, it just happened again, except this time EVERYTHING went black and white, forcing me to shutdown. I'm damn tired of this, and driver updates don't help it. Vista x64 Ultimate is what I'm using with 64x ATI drivers for my 4850. It's pissing me the fuck off, excuse my language, especially having that last one happen.

What is this?
 
I've had this problem too, but it's a very generic problem because you don't know what is actually wrong. I'm guessing your eventlog will say something about atikmdag. It only indicates that "something" has caused a problem with your display driver. It could be your video card being faulty, overheating, lack of power, bad system ram etc, some devices in your pc not working well together etc... There's no one solution for it unfortunately
 
madh83 said:
I've had this problem too, but it's a very generic problem because you don't know what is actually wrong. I'm guessing your eventlog will say something about atikmdag. It only indicates that "something" has caused a problem with your display driver. It could be your video card being faulty, overheating, lack of power, bad system ram etc, some devices in your pc not working well together etc... There's no one solution for it unfortunately
Click to expand...
Well it's not like my system gets too hot, and I've run memtest and all that before...if anything I could guess the card is faulty somehow, but I don't know why it would just come back on if so? I'm pretty sure it's a software problem, but as I said, I don't know.

Wazooty said:
Turn down your in game settings.
Click to expand...

...why would I do that?

Vashypooh said:
What powersupply do you have.
Click to expand...

550W Modular Rocketfish. Did the same on my 700W or whatever.
 
j-sta said:
because that would help determine if it's a software issue, or hardware issue.
It would stress the video card less, helping rule out a power/hardware issue with the VC.
Click to expand...

Well it hasn't done it in more hardware intensive games such as MMOs, yet does it in L4D.
 
Guitarrasdeamor said:
Well it hasn't done it in more hardware intensive games such as MMOs, yet does it in L4D.
Click to expand...

How hardware intensive a game is depends very much on how high your settings in the game are.

I run WoW fully maxed out, but crysis on gamer (high) will crash my vid cards driver, because it's simply too much for a 9800gtx to play at very long, at my resolution anyways. If the MMO you're referring to is wow...it's not that hardware intensive Hardware will improve it, but it's not intensive like crysis is intensive.
 
it's just L4D.......

It's always had problems like this. Like every other driver (ATI or Nvidia) fixes it then breaks it again.
 
Wazooty said:
How hardware intensive a game is depends very much on how high your settings in the game are.

I run WoW fully maxed out, but crysis on gamer (high) will crash my vid cards driver, because it's simply too much for a 9800gtx to play at very long, at my resolution anyways. If the MMO you're referring to is wow...it's not that hardware intensive Hardware will improve it, but it's not intensive like crysis is intensive.
Click to expand...

AoC and CO.

Bigbacon said:
it's just L4D.......

It's always had problems like this. Like every other driver (ATI or Nvidia) fixes it then breaks it again.
Click to expand...

Doesn't explain GTA.
 
Its your video driver crashing. Now how to fix it is another issue. A lot of times drivers crash due to conflicts with other hardware or software. Its rough trying to pinpoint which piece of hardware or software is causing the crash though.
 
Check your VRM temperatures. Are you overclocked?
When i first bought a 4850 they would do that because the VRMs got hot. It depended on the game, sometimes it would be fine other times it would overheat and cause an error.

just check it using GPUz
 
this crap happens with my nvidia based system once in a while while just browsing the web. its all drivers as far as im concerned.
 
Not OC'ed. Saying my GPU temp is 100*C right now..been playing L4D for half an hour or a bit longer maybe? Turned up AA and AF actually to the top and no crash so far...
 
100c is too hot I would say. Find a way to cool that thing down and see if it helps.
 
Ive had this happen on my PC when my partner was playing Sims 3 and had forgotten to speed up the fan from 24%. It would last maybe 5 mins and then the driver would crash out and recover then crash out 5 mins later again. Speeding up the fan stopped this.

HD4870
Core 2 Duo E8400
 
If you haven't already, I'd compress air it and try running it at 75% fan speed.
 
I had exactly the same problem - 4850 in Vista x64, driver crashing and recovering in GTA4 (and Wolverine, and just CTDing in Riddick:AoDA. I think L4D was fine, though...).

Happens with Catalyst 9.4 - 9.7 (haven't tried any newer ones), but 9.3 works perfectly. Give them a try.

Where are your temps at with the fan at 75? Did it fix the problem? 100C is definitely unhealthy...
 
When idle, it was around 60* at 75% I believe. I'll check when I get home. 9.3 worked good? Dust could be a problem...I shall pull it out today and have a looksee. Don't know if I'm comfortable with taking it apart, but couldn't be too hard (just never pulled a video card heatsink off and cleaned it).
 
Guitarrasdeamor said:
Dust could be a problem...I shall pull it out today and have a looksee. Don't know if I'm comfortable with taking it apart, but couldn't be too hard (just never pulled a video card heatsink off and cleaned it).
Click to expand...
60C isn't too bad. I'd still check the load temps, but it sounds like the fan speed was the only thing causing overheating, and I doubt you'll see much difference from dusting it.

If you decide to do it anyway, since you say it's your first time pulling off a heatsink, it's probably worth mentioning that you'll need some thermal paste handy when you put it back on. Otherwise, your temps are going to get a whole lot worse...
 
If your GPU is idling @ 60 and full load @ 100c that would cause your problem.. clean out your pc fix the air flow and up the fan on your gfx card and unless you have already damaged it from heat that should fix the problem.
 
While 60c isn't horrible.... I can't stand that high of temps.. especially at idle.

I take it that it has the stock reference cooler on it.

If you have room in your case.. most likely do, I would look into getting an Accelero S1 v2 to slap on that video card. Even without a fan, your load temps will be lower than your idle temps are now provided you have decent air flow through the case.

And, if you are able to get some air moving over the Accelero you can expect very low load temps. Currently, my 4850 overclocked to 740/1150 gets load temps of 42-43c with the Accelero S1 v2.
Idle temps are 33-35c depending on the room temp.

The drop in GPU temps is definitely worth the small amount of money for a way better cooler - $25 at Newegg.com

http://www.newegg.com/Product/Product.aspx?Item=N82E16835186016
 
I had this problem only it was on my 8800 GTX. I troubleshooted this problem for months. Re-installed windows, tried XP, Vista, Every different driver i could think of. The only thing that stopped it for me was under clocking my video card. Not just using def clocks, but actually going lower. The temps didn't seem to high but during the RMA process with EVGA they recommended i under clock the card 10-25% to see if that stops the crashing and it did. Maybe something you could try it definitely helped confirm that my card was bad. Good Luck
 
turn the fan to 100% and see how hot the card gets. If it cools it down and you can play without the error then you know for sure its a heat issue which it definitely sounds like. Maybe even play with the side off the case as well. If the problem does not reoccur then you should look an an aftermarket cooler for that gpu. what make of 4850 do you have?
 
GPU's are rated for higher temps then CPUs. HD4000 series is safe until about 110c according to ATI. I've seen one of my HD4870s (Crossfire) hit 100c before and the fan kicks in at 100c and drops it back below 100c. It never caused me any artifacting or crashing but I wasn't comfortable with it pushing 100+ so I manually control my fans now.

One GPU idles at about 76c @ 46% fan which is plenty safe and doesn't bother me in the least. Plus I think the sensor is faulty and reporting higher then it really is. The other at about 66c @ 42% fan. When I game I boost the fans to about 70% and that keeps both GPUs between 68-75c with any game I throw at it.
 
obliv said:
I had this problem only it was on my 8800 GTX. I troubleshooted this problem for months. Re-installed windows, tried XP, Vista, Every different driver i could think of. The only thing that stopped it for me was under clocking my video card. Not just using def clocks, but actually going lower. The temps didn't seem to high but during the RMA process with EVGA they recommended i under clock the card 10-25% to see if that stops the crashing and it did. Maybe something you could try it definitely helped confirm that my card was bad. Good Luck
Click to expand...

Ouch... doing multiple windows installs before trying to underclock.

100C seems a little high IMO. I would just see how much dust build up there is, and also probably reapply the heatsink which can drop temps a fair amount depending on how well it was applied stock. There are a few models that have no heatsinks on the VRMs either. Mine was like that, I slapped some on it before installing as some people said they seen artifacting and solved it by getting some vrm sinks.
 
Last edited:
Vermillion said:
GPU's are rated for higher temps then CPUs. HD4000 series is safe until about 110c according to ATI.
Click to expand...
Well, "safe" as in safe from catastrophic failure. Not necessarily immune to crashing and/or artifacting. Though it's likely a conservative estimate.

But note that the OP said
"Saying my GPU temp is 100*C right now..been playing L4D for half an hour or a bit longer maybe?".
I don't know about you, but in the time it takes to quit a game and open a browser (3-4 seconds, maybe), I'm just about back to idle temps. Unless he likes running FurMark while he's browsing [H], I'd guess that he was idling pretty close to 100C. Which is definitely cause for concern.

Sounds like it's under control now, though... Just the 4850's shitty default fan control on top of its already sub-par reference cooler. (Mine came with a Zalman attached :p . Cheapest 4850 they had, too... :confused:)
 
Tbird87 said:
Ouch... doing multiple windows installs before trying to underclock.
Click to expand...
Well, it doesn't exactly look like a hardware issue. When I think hardware issues, I think bluescreens, freezes and artifacts. A three-second pause and a tray notification about a driver problem, and I'm more likely to suspect the drivers (or some other piece of software that's not getting along with them).

And in my case, I turned out to be right... Happened every five minutes with Cat 9.4-9.7, and not once in hundreds of hours with Cat 9.3.
 
Well I'll try some stuff like cleaning out and temp/fan control today.
Should the fan on the video card be blowing out/in? I thought I had both my case fans on the sides wrong when I put together my case when I put it together, but never changed them...those two 120mm fans I have on the side might be placed wrong :confused:

LuminaryJanitor said:
Well, "safe" as in safe from catastrophic failure. Not necessarily immune to crashing and/or artifacting. Though it's likely a conservative estimate.

But note that the OP said
"Saying my GPU temp is 100*C right now..been playing L4D for half an hour or a bit longer maybe?".
I don't know about you, but in the time it takes to quit a game and open a browser (3-4 seconds, maybe), I'm just about back to idle temps. Unless he likes running FurMark while he's browsing [H], I'd guess that he was idling pretty close to 100C. Which is definitely cause for concern.

Sounds like it's under control now, though... Just the 4850's shitty default fan control on top of its already sub-par reference cooler. (Mine came with a Zalman attached :p . Cheapest 4850 they had, too... :confused:)
Click to expand...
I was running L4D while browsing the net when I died on infected. :p what I always do.

Also my card is a Diamond.
 
LuminaryJanitor said:
Well, it doesn't exactly look like a hardware issue. When I think hardware issues, I think bluescreens, freezes and artifacts. A three-second pause and a tray notification about a driver problem, and I'm more likely to suspect the drivers (or some other piece of software that's not getting along with them).

And in my case, I turned out to be right... Happened every five minutes with Cat 9.4-9.7, and not once in hundreds of hours with Cat 9.3.
Click to expand...

I didn't quote your situation :confused:. If it looks like software, troubleshoot the main things you think are the problem, but theres a certain point where you can try other little things before multiple os installs. My point was about doing the tasks that take a few seconds to do verses multiple os installs. Maybe that's just my preference on how to troubleshoot. It wasn't about some argument about it being software or hardware. That statement really wasn't pointed to the OP. Maybe it looked that way to you, didn't to me.

I'm just adding on additional things to do. When my 4830 was giving me problems, it gave me a wide variety of quirky shit. It worked better with the memory underclocked and seemed fine for a little while. Then it just died hard. If my video card was hitting 100C, having problems or not, I would still recommend doing what I stated. I was just throwing in anything else that could be a factor.
 
Last edited:
Still idling at 68* right now with fan at 70%. With it at constant 100 it seemed to do it at 60*...I took it out and blew it off with air as well.
 
Guitarrasdeamor said:
When idle, it was around 60* at 75% I believe. I'll check when I get home. 9.3 worked good? Dust could be a problem...I shall pull it out today and have a looksee. Don't know if I'm comfortable with taking it apart, but couldn't be too hard (just never pulled a video card heatsink off and cleaned it).
Click to expand...

For comparison, my 4850 is idling at ~51C at 75% fan speed. I haven't blown it out or anything since I got it a year or so ago. I don't have any artifacts or driver crashing problems.
 
scr1251858419.jpg


i think my gpu hits maybe 80*C under load, but again i have pretty good airflow in my case..

edit: heres one after a few mins of RTHDRL

scr1251858585.jpg
 
You must log in or register to reply here.
Back
Top