Weird artifacting

Technology and Gaming Help and Support
#1

This morning I noticed some weird artifacting on my computer. It happens usually on a white background (like that of a window) but when it gets more extreme it goes anywhere. The first time it did it I assumed it was my monitor (which is fairly old) but when I tried a different monitor it still did it. With the 2nd monitor (although not because of) it gave me the “display driver has stopped working” message and then BSOD’ed telling me that the problem was the nvidia driver.

I took the graphics card out and cleaned some of the dust out of it (it is quite dusty but I haven’t got any canned air at the moment, getting some this week probably). I waited until the graphics card was back to room temperature and put it back in the case and turned the pc back on. I used the pc for a while checking the temperature of the graphics card every so often and it doesn’t go over 60 C which seems fairly acceptable to me. I triggered the artifacting to happen again by opening photoshop and editing a photo so I assumed it was open GL related (photoshop cs4 uses open GL to render stuff). I shut it down before it had the chance to BSOD, left it for half an hour or so and booted up again.

I then installed the latest nvidia drivers (I was a few months out of date) and the problem didn’t occur again for ages (couple of hours). When it did come back on it was very localised (only the box you get when you press the start button) and didn’t crash so I took a screenshot of it.

It now seems to show up and dissapear completely randomly although other than being visible it doesn’t seem to cause any problems. The times it’s been worse there has been multicoloured dots all over the screen and then a kind of aura of dots following my mouse pointer around. My graphics card is still in warranty and such I just need to find the invoice for it. Is it likely to be a hardware problem or is it something software related as I’d rather be able to fix it at home than be without a computer for as long as it takes to get it repaired or replaced.

Edit: Computer specs are as follows
Intel Core 2 Duo E8300 @ 2.86 GHz
2GB Corsair TwinX 800 MHz
Asus P5Q
EVGA 8800 GT 512MB
Vista Home Premium

I built the PC myself last year and this is the first problem I’ve had with it. A friend of mine who works with PCs suggested that it could be a memory problem on the graphics card.

My brother has an identical machine (built the same PC for him as well) and so I can test the graphics card in his PC but as I haven’t got a way to make it happen this test could prove completely useless, the artifacting happens completely randomly and either goes away seconds later or crashes the PC.

#2

I had a card that got overheaded once to often, and would produce artifacts like those in windows, and in games, tho in games it was either a “funky pixel” on the screen, a messed up texture, or messed up geometry or model, or combinations of the above, and as time went on (e.g. in Crysis) it’d get worse.
The thing is, once damaged (presumably by overheating in my case), it’s not going to get better. It just gradually dies till one day it’s completely dead.

#3

The thing that strikes me as odd is that the card hasn’t been overheating (or at least not to my knowledge). When I say that I let it cool down, I simply mean back to room temperature from it’s idling temperature. I don’t play paticularly intensive games (TF2 mostly) and I’ve never seen it go higher than 80-90 C. It’s not overclocked and my case has good airflow. That’s what makes me think it could be software related rather than the card randomly developing a major problem for no reason whatsoever.

If it carries on, once I find my invoice then I’ll RMA it if need be but I just don’t want to go through the trouble of an RMA only to get back a new card and still have the same problem because it’s software related.

#4

You mentioned cleaning out the dust, so at some point past it may have overheated, causing damage. Thus if that is the case, it’s already broken.

#5

Yeah, i’d go with the software issue. Clean the original drivers off, then download the latest ones. I have the same card as you and i have never had a problem like this.

#6

Okay, sounds like a plan. Do I just uninstall them as though they were any other piece of software or is there something else I need to do?

#7

Yeah pretty much. If you wanna go extreme, some people use that Driver Cleaner thing after uninstalling the driver, but I don’t think that’s really necessary.

Basically, you DL the newest driver, go to Add/Remove Programs, search for NVidia Drivers, uninstall, reboot, install the new driver, reboot again.

#8

It went majorly fubar this morning and so I uninstalled the graphics drivers and other nvidia related stuff (physx etc) in safe mode. It started up again and installed some nvidia drivers automatically. It looked like it was going to be OK for a few minutes but then the problem came back a few minutes later.

Although the multicoloured spots seem to come up and dissapear with normal use, it only seems to blue screen/crash and burn where Open GL is concerned. This is problematic because firstly I take and edit a lot of photographs, the editing can wait so that’s not that big a deal but the process of an RMA could mean that I have no more storage space for photos on my camera. The second problem is that my screensaver is rendered in Open GL. This may sound like a non issue but the problem lies here, my screensaver comes on after 10 minutes. Pretty much, if I leave the PC to do anything, the screensaver is going to come on and the PC will blue screen. Secondly, I can’t turn the screensaver off because going onto the screensaver menu thing on vista means that it will show the little preview box which, you guessed it, starts rendering in open GL.

Sadly, it looks like I’m going to have to RMA the graphics card. Thanks for the help anyway guys.

#9

You’re gonna need a new graphics card soon. I’ve had that problem with a geforce 4 and a geforce 5200.

Single slot graphics cards overheat more often then dual-slot cards due to smaller cooler.

#10

Yeah it sounds like a borked card. How is it handling games at the moment in DirectX?

#11

The computer is basically unusable now anyway. I haven’t tried a Direct X application on it but in general windows use it still has artifacting it’s just that Open GL seems to make it crash. The place I bought it from (Scan which is one of the biggest computer hardware e-tailers in the UK) is right near my house (their HQ/warehouse) and they have a returns/tech support thing there. I’m going tomorow with all my warranty info/invoices and hopefully they’ll refund/exchange it in the store rather than me having to RMA it. The card itself has a 2 year warranty but Scan (as far as I remember) give you a 1 year warranty after the date of purchase. Ironically, it’s 13 months (almost to the day) that I bought it but often the guys there are open to negotiation and I’m hoping that in the end they will just exchange it in store as they won’t lose out on anything in the end anyway.

What strikes me as odd is that this problem has occured from literally nothing. Although the card is dusty it’s not overheated, I always take precautions to discharge static etc and I know how to handle a PCB without breaking it as I’ve been doing this for quite a while now. It’s not even like I’ve overclocked it or anything. My last home built PC went under a lot more stress than this one and finally burnt itself out when I damaged the power supply. Also due to my computer not having onboard graphics, it is rendered completely useless to me until I have a replacement.

#12

There could be thousands of reasons why it died, most likely the card was destined to die soon anyway.

#13

Yeah, sounds like it could be faulty. Unlucky = [

#14

Ah well I’m going to the place today to battle it out with customer support. Hopefully they’ll let me exchange it rather than me having to RMA it. Sometimes they’re quite reasonable, let’s hope one of those times is today.

#15

Update: Just got back from scan. Was there for about half an hour. The tech support guy went off and tested it and confirmed that the card was fubar. He offered me an £80 refund which I was gonna take but it turned out that the card was 1 month out of their warranty so they could either send it off for an RMA or trade it for a “similar item”. I asked him to find out what “similar items” they had and he came back with an 8800 Ultra.

Lopez 1 - World 0

#16

Win

#17

major win

#18

My geforce 6800XT overheated, it had those same light blue pixels show up on screen, my motherboard’s northbridge heatsink even fell off because of the heat, it was an old pc that i was about to replace by this one.

#19

Alas the problem seems to have returned with the replacement graphics card. I took it back this afternoon along with my motherboard in order to check if the motherboard was the issue. I/they couldn’t replicate the issue using my motherboard but the replacement graphics card was completely wrecked (even worse than the original, can’t even boot to windows now). I’ve sent that one off for RMA and the tech guys in the store assure me that it’s not a motherboard issue. I’m currently using a replacement that I bought as they had no other cards they could offer me as a warranty replacement in stock. I’m going to use this one until I get my graphics card back from the RMA. Let’s hope this one doesn’t suffer the same fate as the other two, if it does, it’s looking like something more serious.

#20

I would keep that mobo if I were you, I keep all of my “killer” hardware, I have a hard drive that can kill motherboards, I have a CD drive that kills CD’s and I have a monitor that kills graphic cards over time.

Not to mention my killer 300w power box.

Founded in 2004, Leakfree.org became one of the first online communities dedicated to Valve’s Source engine development. It is more famously known for the formation of Black Mesa: Source under the 'Leakfree Modification Team' handle in September 2004.