Freezing - GPU hardware faulty?

[xxxlm]

Hello, I'm experiencing a regular and specific problem when playing games like Crysis and Bad Company 2:

5-10 seconds after having loaded a map in either game, the screen freezes and a red pattern of pixels blends into the screen. Then either Windows shuts the game down, or becomes non-responsive -> hard reboot.

The GPU temperature shouldn't be able to rise that quickly after just a few seconds, so I don't think it overheats.

My system:
Asrock X58 Extreme3
I7 930 2.8Ghz
A-Data 3-channel RAM 6GB
Gigabyte GTX 480
HDD and DVDRW
Cooler Master HAF 932
950W Corsair PSU
Latest GPU drivers
Win 7 64 bit

Nothing is overclocked, the chassis is standing on the floor in a normal room temp setting.


So I'm just wondering, is this kind of problem a symptom of a single badly manufactured graphics card? Or could it be some chip on the motherboard that doesn't do its job?

I can play other games like Starcraft 2 at the highest settings without crashes. The computer is barely a week old and Crysis/BC2 worked flawlessly until a few days ago. Thinking about whether to RMA the GTX 480 or maybe change motherboard. Help =)

 

[skolpo]

Even though it doesn't crash on Starcraft 2, it can still be a graphics card issue. Do you have another card to test with the system? Or perhaps another system to test the GTX 480 in? Whatever you can do to isolate any type of hardware you think that might be causing the problem, the quicker you'll find where the problem is coming from. Everything else on your computer seems more than efficient enough to run the games with maximum stability.

One thing I'm wondering is if the power connectors (6-pin and 8-pin PCI-e connectors) are connected correctly onto the GTX 480. There might be a lack of power running to it if the connectors are not properly connected. Since Starcraft 2 isn't as demanding as Crysis of BFBC2, you might not be able to encounter the same issues in the game if it was a power connector issue.

Another thing you can check is your graphics drivers. Do a clean sweep of your graphics drivers and try reinstalling it again.

 

[xxxlm]

Thanks for the input skolpo,

I've checked the 6- and 8-pin connectors, the 2-pin part of the 8-pin may have been a little loose, but I don't think it was. Also did a clean sweep of the Nvidia display drivers (with Driver Sweeper) then reinstalling the latest ones.

Multiple problems still there though, check this out:

BIOS loading freezes
- A pattern of blue colors when loading Windows -> Unresponsive
- Black screen when loading Windows -> Unresponsive
Windows freezes
- Laggy mouse icon followed by freeze in Windows (no weird colors) -> Unresponsive
- Regular freeze in Windows -> Unresponsive
- Watched Youtube video, screen freezes, SOUND LOOP -> Unresponsive

And a few game freezes as well, with red or blue color patterns on the screen. Sometimes with sound looping, sometimes not.

I've used HWmonitor to check GPU temps, typically they're at 75-85 celsius when playing games. And since most freezes do not occur when gaming, I don't think this is the problem.

Updated BIOS to P2.1 (latest from Asrock).


Having restarted the PC after a freeze, the BIOS clock resets itself to Aug 24 for some reason. Not sure if this is relevant.


What's weird is: I've been able to play Metro 2033 in 3D Vision on high settings without any issues, same with Just Cause 2.
I've played Mass Effect 2 on highest settings for up to 5 hours without a problem, shut down the PC properly, started it up an hour or 2 later -> BIOS freezes.

Sometimes BIOS freezes multiple times, and the only way get it working is to boot in Safe Mode, then restarting.


skolpo, or anyone with experience, does this sound familiar in any way? Power supply not working? Motherboard not working?
Defect GPU? I'm getting another graphics card to test with next week, if that doesn't work, I'm not sure what part to RMA

 

[EXT64]

Is this a completely new system? In other words, are there no parts we can rule out (since they were previously working)?

 

[skolpo]

I'm starting to think that it's a funky GPU. Try a new card and see if the problem persists.

 

[xxxlm]

Yeah, it's brand new, had it for 10-12 days now. And the problems started about 3-4 days after I first started using it. Anyway, will borrow another graphics card in about 3 days so I'll try that first.

 

[Kari]

while you are waiting for the other card to arrive, check the RAM with memtest86+. Let it run for multiple cycles, ie several hours. if you get errors, run it again with just one stick to find out which one is faulty. (or remove one each time..) memtest86+ is a small bootable program, you can get it from here
http://www.memtest.org/

I was getting similar pretty random crashes with funky colors and looping sounds when one of my ram modules was faulty. Some games would crash almost instantly while others run fine for several hours, windows was stable most of the time but would sometimes crash on a seemingly trivial task, like opening web browser etc. In the end while i was troubleshooting the problem, and booted up with just the faulty ram installed, it froze while loading up windows and in the same time messed up the pagefile.sys file, causing it to BSOD everytime windows started even with only the good rams installed. Booted into safe mode, deleted the hidden system file pagefile.sys and problem was solved. (windows will recreate the file if not found, so no worries)

 

[xxxlm]

Hmm, any idea where the pagefile.sys is located? I'm still a newb at Windows 7

 

[Kari]

C root, but i think it was just very bad luck it got corrupted in the process.
And if you only boot straight into memtest it shouldn't mess with windows even if the ram is faulty. I was just curious and didn't know faulty ram could cause something like that to happen.

Testing the ram with memtest should be completely safe, so don't let my unfortunate experience keep you from troubleshooting the problem.

 

[nforce4max]

RMA