Consistent Graphic Glitch

I'm getting a new problem with my laptop now, specifically with the graphics card. Apparently, it doesn't matter what I do, at some points whether I'm playing a game or not (but it happens more often during a game), the screen will suddenly become broken and segmented into fluctuating black and colored lines, almost like the static you'd see on a TV channel you don't get. When I'm in a game, I can usually stop it by minimizing the game and re-maximizing it, but it gets annoying constantly getting this problem and fixing it this way. I've tried both the 93.71 drivers and the 93.81 drivers, but the problem persists regardless, so it's not a driver problem. It happens with games as old as Homeworld: Cataclysm all the way to new games like S.T.A.L.K.E.R. I'm thinking it may be a physical problem that's probably going to have to get the card sent away for repair, but I want to see if anyone knows what's happening and how to stop it.

 

Could be artifacts caused be heat.
Have you tried running your card at stock clocks and see if it still happens.

or at your OC stock check for artifacts with ati tool.
http://www.techpowerup.com/atitool/

 

Right now, the clock is set to the lowest setting that I can set it, given the .ROM files that are available. It still happens.

 

What kind of temps do you get while gaming at lowest clocks.

 

If I set the fans to High speed with no supportive air flow, it can get anywhere from 70 to 80 degrees Celsius. But, if I use an external cooling system (right now, it's just a table fan blowing across the desk surface to get under the laptop while it's propped up at the back, but I am getting the Spire Pacific Breeze soon enough), it gets to about 60 to 65 degrees Celsius. In fact, in order to prevent MFI Parity Check Errors happening, I need another cooling system, which is the only way I can play S.T.A.L.K.E.R.

Thing is, with or without an external cooling system, the glitch still occurs.

 

You might have a badly mounted heatsink. Call Dell ASAP, and get them to fix it. As you have overclocked it, you also may have caused irreversible damage to it, or the RAM is damaged due to heat (it's not measured). Happened to my old M6805.

 

Well, I did OC it, but it was no more than the "safe" number that was stated in these very forums. And I don't think it's the RAM, because memtest86 and prime5 both saw no problems with it.

I do believe, however, that it could be the heatsink. Recently many games have problems playing because of either MFI Parity Check Errors or these glitches I mentioned, and I do notice that the GPU temperature is well into the high 70 degrees when I have to reboot. I think Dell might have to replace it altogether.

 

Just because it's considered safe doesn't mean that it is. It still generates more heat, and if your system was marginal at the basic speed, the little bit of overclocking may have pushed it over the edge. Call Dell ASAP, because otherwise it'll just get worse, especially since it's now doing the same stuff underclocked. That's a sign of actual damage, not just temporary overheating.

I wasn't talking about system RAM either. I was talking about your graphics memory. You know, the 256 unshared (really should be called "dedicated") you mention in your sig. That is changed with the bus speed, and it is subject to overheating and damage as easily, if not more so, than the GPU itself, because it rarely has it's own heatsink.

 

Hmm. I figured as much. Kinda sucks, since it's going to take weeks to send, fix and get back from Dell...

 

Sorry bud. I wish I wasn't the bearer of bad news. Back up what you can, as they have a habit of hosing drives when they "repair" them. Best bet is a hard drive image to an external drive, with Norton Ghost, or if you feel geeky, a Linux bootable CD and dd.

 

I already backed up the needed data. Rest is just programs that I can install later on. If they feel the need to reinstall the OS, they can, I can just reinstall the programs when I get it back. Although, seeing as how they're just replacing the GPU, it shouldn't require a reinstall of the OS.

 

I just spoke with Dell Support. The person I spoke with, after explaining the problems, told me to try and hook up the computer to an external monitor. So I did, and switched back and forth. I told them that it wouldn't matter, because the problems were random in occurrence. So, they told me to update my BIOS from A06 to A07 and to do my normal stuff in case the problems occur again, and to call back.

Not a very helpful support session.

On a side note, I updated my drivers from 93.81 to 97.92, and I've actually noticed that those graphical glitches that were happening have been significantly reduced. Where once I get this happening quite often, now it only happens once or twice. Although it would be nice if the problem would go away entirely...

I think I'll just get them to take it and replace the card. I'll also see if they can put in a dual heat pipe system instead of a single pipe system, to reduce the temperature. They most likely won't, but it wouldn't hurt to ask.