Suppose your desktop looks like this

[The Brick]

lol used video cards lol

It's new.

 

[Rico]

My guesses:

1) Your case is getting too hot and your videocards can't handle it.
2) PSU may not be supplying enough juice to the card
3) Driver problem? Windows problem?
4) Computers hate you.

Check your case temperature to make sure its not too hot, then try reinstalling your drivers. If that all fails, try borrowing a PSU from a friend and replace the one on your PC with his, to see if thats the issue. I'd only reformat and reinstall windows if really necessary.

 

[Hectic Glenn]

Try underclocking it.

 

[The Brick]

Temps are absolutely fine. Doubt my 550W PSU could handle an 8800gtx but not a much more power efficient 9800GTX+. Can't imagine the drivers being an issue. Ran on 185.20 when it crashed last week, went to 181.20 WHQL and it crashed exactly the same yesterday. Always clean driver installs. Doubt windows is doing it (no blue screens). But yes, computers hate me at times.

Weird thing is, I've been running heavy games all week, for long periods of time, yet both crashes were when I was just nothinin' about in windows.

 

[Rico]

It might be a good idea to tell us exactly when the crashes happen (are they really crashes? They seem like artifacts to me). What temperatures is your video card reporting? It still seems like artifacting caused by a bad video card to me. Are you sure the new video card isn't refurbished from the factory?

EDIT: If you have any friends with video cards, ask them if you can borrow theirs for a bit and swap yours out for theirs. See if the problem keeps happening. If it does, it's not the videocard. If it is, then you found the culprit yet again.

 

[The Brick]

Before I got the replacement card, I had been running perfectly fine on a friend's 7900gtx. Temps are 35?C/64?C, nothing this thing couldn't handle. The packaging it came with was flawless, not sure how to determine the newness. It was completely dust free.

And they are very much crashes. I'm just doing nothing. The first crash (last week) was when I clicked to launch xfire. The pc instantly froze completely. There's no buildup of artifacts, I haven't seen any artifact yet. The second time it crashed, I switched a song in Spotify. It instantly froze again with a garbled screen almost exactly like the first one. But this time, the pc rebooted after 10 seconds or so. If this happens once more, I'm gonna try to get it replaced... sigh.

 

[Rico]

If your friends videocard still works well on your PC, I'd say you got another bad card. Were it to be something else, the problem wouldn't really go away when you switched out cards unless your PSU couldn't handle the new card.

Even though the card looks new, it could still be refurbished. Basically refurb cards are returns from customers, or a card that initially had defects but was repaired by the manufacturer and repackaged. They are often sold at a discount for obvious reasons, but sometimes manufacturers will issue them as replacement cards.

Just to be sure, would you happen to know the Amperage going through your 12V rail? Should be something around 26A. Wattage isn't a good way to gauge PSUs as lower end manufacturers like to claim high wattage but the PSU ends up using that power in all the wrong places, often times causing people headaches because they have a 1k watt PSU that they think should be able to run everything, but isn't supplying enough juice to their most power hungry components.

 

[The Brick]

http://www.antec.com/usa/productDetails.php?lan=us&id=28550

Doubt it's causing trouble. The crashes happen when it's idling.

 

[Rico]

Then I'd try a new card, as it is the most likely culprit. Other than that, other than a faulty motherboard, I don't see what could be causing it.

 

[The Brick]

aaaaaand crash number three.

http://i208.photobucket.com/albums/bb287/tharbrick/fsdfsd001.jpg

Right, gonna ask for a replacement again. They didn't respond to the email I sent before. Sigh..

Edit: and email sent:

Hello,

I hate to have to come back to this, but this replacement card is crashing every so often, every time in exactly the same way. It crashed for the third time just now. Here's a photo of it.

http://i208.photobucket.com/albums/bb287/tharbrick/fsdfsd001.jpg

This is exactly like the first crash I emailed about (quoted below, I haven't received a reply). The second crash was also the same, but the system rebooted after a few seconds, so I did not have to time to take a photo. All of these are under idle conditions. Temperatures are fine (34?C idle, 64?C load). There's no signs of instability other than these rare crashes. I'm using WHQL drivers (181.20), and switching to other drivers (185.20) did not affect the problem.

I'm hereby requesting a new card. Will I need to open a new RMA request, or can this be handled through email?

While this card is away, I'll be using an 7900gtx again. If such a crash occurs again with a different graphics card, I will notify you immediately and pay for the shipping to have my card returned.

Any reply is appreciated.

Regards,
Daan

 

[The Brick]

I got a nice response this time and I will be sending my card back tomorrow. They are going to test the new card before they send it to me.

 

[Vegeta897]

Why would you want it replaced, your video card turns your desktop into a game of breakout. That's awesome. I wish I could play breakout all day instead of doing work.

 

[morgs]

My brothers 8800 just crapped out too.

 

[Souseiseki]

I've had this happen on a FX5200 ages back (2004) and the card is very cool (no, the card sucks in fact, but the temperature). I don't remember exactly what causes it but forcing Vsync on absolutely everything did cause issues and cause the card to freak out (for me at least. I know the desktop doesn't use 3d accelleration but this was a factor in my issue).

If you're desperate for a non-glitched desktop, just suspend, or hibernate, then come back and it'll be unglitched. It's only a workaround though. I have no idea how it is on the newer geforce models.

I use Radeon now though, which is less forgiving than what you get:

 

[The Brick]

No, these cards were really broken, the system would completely crash. Read the thread.

 

[MuToiD_MaN]

Brick, I'm going to use your screenshot and make a new tile background for my Windows desktop. That thing's awesome.

 

[The Brick]

First, goddamn ****piss cockshit.

Right. Three months later.. I figured out what the problem was. Story cut short, these cards simply fail to run stably at their stock speeds. It's silly and fail of me not to have tested this earlier, but running an 'artifact scan' in ATItool recently (after crashing problems with red alert 3) made it pretty obvious that this card was utterly unstable. It would give deltas of thousands of pixels within a minute. After bringing down the core clock from 740mhz to 700mhz, it would run perfectly fine for over 30 minutes.

I'm delighted that all the problems are gone, but I'm pissed off about how it just fails to run properly at stock speeds. Such a silly issue that caused me so much trouble (on my 3rd card now).

Big sigh.

 

[MuToiD_MaN]

You had me at "goddamn ****piss cockshit."

Good story, man. And remind me never to buy a card by the name of "sparkle."