*Help* Graphics card broken?

I’m really in need of some help here.

My PC locks up when logging into windows unless i run i “safe-VGA” mode. It will install the forceware drivers, but hangs after that. We’re talking total lock, no BSOD or error messages. I’ve tried to remove evrything else (network, disable sound, tried 2 different HDDs) but only thing that makes it stable is running in VGA.

Two days ago we had a couple of power-outages - could that be the cause?

System : Opteron175 (no OC), GF6600GT, WD Raptor HDD

Oh on a side-note - wiondows displays the HDD as a removable device under “safe-removal” icon WTF is up with that - is it usual for SATA-drives?

Any insight will be very appreciated!


its normal for sata drives, you can hide it somewhere (dont recall)

as for the card: is the fan spinning? if not, replacing the fan is pretty simple and easy. I’m pretty sure if it is something else though, the card isnt repairable.

Fan is running ok.

What I’ve done now is I’ve moved the card to the other PCIe slot and reinstalled the drivers. Still getting random lockups when running in normal windows, seems to be stable in safemode. Seems only option short of getting a new card is trying a clean windows install (universal solution, right?).

Really, really pissed, since the system is only about a week old, and I’ve just started playing WoW (which is even better than I thought it would be…)


I’d try the card in another mobo as well- it might be the mobo, which would suck to find out after buying another 6800. Might be instead ram I guess, so try different ram / try your ram in another pc. The power dealie could definately have done something, especially if there isnt a surge protector or UPS or anything in between the pc and the wall outlet. I think its mroe an issue of when the power comes back on than it going out. Also ‘dirty’ power can kill stuff in odd ways, but your earlier pcs would also have been weird. I think.

Anyone else going to help this guy? I’m fumbling!

I’d honestly say reinstall windows. Reboot after the chipset + graphic driver installs. Install one game, and see how the sucker runs.

If there’s no crashes, then you’re all cool.

Possible it’s a corrupt file system. It’s just so much easier if it’s a software problem (only have to reinstall).

If you still crash with a clean install then check the usual suspects

  1. run memt86 off a cd overnight.
  2. Unplug the power, reseat all your RAM and video card.

To be honest if you say you run stable in safe mode it sounds like software/driver problems. Which is good, cause means no hardware is broken.

One thing to do once its up is pay attention to the CPU tempature. Running a graphic intensive game can cause it to get very hot and if the fan isn’t cooling very well, it may overheat and lockup.

So I tried reinstalling windows, everything went fine and I started WoW (need the fix, dammit), After 15-20 I get lockup and reboot. F*ck!

I can’t try other RAM or mobo, don’t have any friends with new enough PC:s. Guess I could take my PC to a service center for trouble shooting, but that would cost a lot (I think the rate is 690 NOK/hour, about $100).

bangs head on keyboard

CPU temp seems ok at 38 celcius, PWM temp is at 45.

I’m really at a loss here - guess I’ll be buying a new graphics card, not sure if I should get a reeealy cheap one (in case it’s not the cause of errors) or go all the way and buy a better one than the 6600. What do you guys think?


$100 is still cheaper than buying stuff when you don’t know what the problem is. I’d take it to the tech if you don’t have a knowledgeable friend who will do it for free.

You never answered the surge protector question. If you don’t have one, now you know why you should. If you do, it’s probably unrelated to the power issues.

I fried a couple modems because I didn’t have them plugged into a surge protector during a storm. Luckily the modem was the only thing that got fucked up.

Oh and whenever I see a thread talking about random reboots and lockups with 939 systems, I offer this:

Try a different keyboard.

I’ve had this happen to two of my friends, though they were both using the MSI K8N Neo. I highly doubt it’s your problem, but you never know.

edit - Nevermind I didn’t see that you said it locked up every time you logged in.

I have a 6 socket outlet from APC which I got bundled with my former Dell PC. Guess it’s a surge protector?

Lockups are varying in time, from almost instant to max 30 min. Guess they could be heat related. The graphics card doesn’t feel very hot, but I’m not sure what’s normal (remember my first GeForce (256 I think it was), you could fry eggs on that sucker.

I’ll try with another keyboard - you never know. Random lockups are such FUN, aren’t they.


I’d actually lean more towards the power supply than the graphics card on this one… Power supply problems can cause lockups.

power supplies are also cheaper than 6600s, heh. AND you might be able to try a friends for diagnoses purposes.

Relax your RAM timings. Seriously, I had a box at work that we built for a guy like two weeks ago and it was locking up, crashing to desktop, spontaneously rebooting and other nice tricks. Took his ram from 2,2,3,5 to 3,3,3,8 and it’s been phat city ever since.

RAM timings are at default, which is - indeed - 3,3,3,8 (1T)
Have tested another keyboard - got crash.

One really curios thing is that lockups come more seldom now - at first it was during the first minute, now I’ve been online for about 30 minutes and still no lockup. Very, VERY weird… I’ll tempt fate with some WoW :-)

EDIT (later, the same day) : WoW crashed after ca 20 min of play. At least I leveled once - heh.

Have stressed CPU+RAM with Prime95 on both cores for about 90 min. no crash - temps are quite ok (38, 42, 41 for CPU, PWMIC, Chipset).
Downloaded 3Dmark, have tried to run it 3 times, it never makes it to firefly forest (2nd test).
So, I guess I’ll be getting a new graphics card afterall, question will be which one.
I’ll sleep on it abd go into negotiations with my wife tomorrow ;-)

Thanks to anyone who has followed as far as this.



Ordered a dirtcheap GF6200 to use in troubleshooting.
Really wanted a better one, but since I’m not 100% sure it’s the GPU thats defective I can’t afford to spend more money.

If it doesn’t work I’ll try another powersupply as DennyA suggested.

Getting a LOT of grey hair and not much sleep over this :-/


Wait a sec. What motherboard is this? Because there are some poorly designed motherboards out there that can’t feed enough voltage to video cards during 3D apps. That would cause the sort of crashes you’ve experienced.

One thing you can try is to increase the AGP voltage to the video card. Yes, it will run hotter, but it is a quick way to check if the crashes are power-related.

I’d lean towards the power supply being the problem. Dropping the power on a supply can still cook it if it was pushing a lot of watts at the time of the failure. Since you’ve already ruled out heat issues, this seems like the most likely case. Especially since it’s killing the rest of your computer and not just blacking out the screen or whatever.


The motherboard is DFI lanparty (came highly recommended…) - powersupply is a Fortron/Source 600W.

I just tried underclocking the memory to 333 - hadn’t thought about that before- will test stability right away.

Well, that didn’t help one bit, if I can run prime95 stable, CPU and RAM should be ok. Was kinda stupid hoping it would help.

Is there anyway to check the PSU in software - smartguardian reports the different voltages - but is that info useful (for example ATX 5.0V is only 4.83, but is that relevant to me???)


I’ll check if I can find a way to increase voltage to video card in Bios - thanks for the suggestion!


If any voltage is substantially off from its rated value, that’s a sign of a power problem somewhere – I diagnosed a power supply problem last year using software voltage readouts.

I’m not familiar with smartguardian – I used Speedfan to find my power supply problems. If you use Speedfan, you want to check the accuracy of the positive voltages (12/5/3.3); the negative ones can be off. I can’t remember how substantial the difference was on my wife’s system was when I found the problem. Right now, both our systems are running voltages within 2% of rated specs.

For what its worth, I agree with others that this appears to be power related.