Bad PSU or bad GPU?

ShaneS429

Distinguished
Jan 18, 2010
3
0
18,510
Okay so I am new to the boards here at Tomshardware but I come here quite a bit.

I build all my computers now. I find it easier and it comes in handy when it is time to upgrade or replace parts. My current build is:

CPU: Intel Core 2 Quad 6600 2.4GHz G0 stepping
GPU: XFX 8800GTS 640MB XXX
Mobo: EVGA 750i sli FTW
RAM: 4x1GB Corsair DDR2 800
HD: WD 6400AAKS & 3200AAKS
1 DVD burner & 1 CD Burner
PSU: 500W Antec

Now the problem I am having just started within the past week. Since I built this machine nearly 2 years ago, I have had no problems with playing games, video editing, 3D modelling, etc. I typically play Crysis, Team Fortress 2, Demigod, WoW, Leauge of Legends, & Unreal Tournament 3.

As you can see, these games range from pretty graphic intensive to not so such.

Recently, I got into the Star Trek Online MMO Beta. I was able to play just fine for the first day or so. Now, I can't even start the game without the computer crashing. Upon reaching the first loading screens, the system either freezes and I hear a feint click sound and I have to manually reset or the PC simply resets itself.

On the Beta forums for the game, other people are having issues similar to mine with the PC freezing or resetting. It was recommended to check temperatures as it might be an overheating issue. I have SpeedFan installed and right now my CPU is running at 30C or so and my GPU is roughly at 57C with nothing open other than Firefox as I type this post.

When I could get Star Trek Online to work, I would alt-tab out and my GPU temp would rise to the high 70s and low 80s. Not enough to cause the system to reset itself I think.

The thing is that ever since I had Star Trek installed and it reset my PC, now any game I play causes the same problem. If I start Unreal Tournament 3, I can start a match, and the PC will reset in about 60seconds. Any game that I use to play before with no problem, now causes my system to freeze or force reset.

I then downloaded FurMark to try and stress test my GPU but once I click the run button, the system resets or freezes. I opened up my PC to clean it out and whatnot, and upon starting it back up, within 10-15sec, I could hear the PC reset itself and the LEDs in my system would get dim. This makes me believe it could be a PSU issue, perhaps it is just going bad, or it isn't enough wattage. So whenever I start a game or some other graphics intensive program, the PC is trying to draw more power than it can handle, and forces a reset. For some reason, the PC starts up fine, but I still cannot attempt to run any type of game.

I have no idea if this is a case of my GPU going out or my PSU going out. I want to make sure I get the right one because I don't want to buy one, and find out the other part is the bad one. I don't have any other PSU or GPUs to swap out and test either.

With FurMark causing the system to reset, it makes it hard for me to figure out what is failing. It could be the GPU failing, so it causes a reset, or it could be that once I start the test, it draws more power than what is available, thus causing a reset. Any help would be appreciated.

Sorry for the long post as well.
 

ShaneS429

Distinguished
Jan 18, 2010
3
0
18,510
Update:

Since I am leaning towards a PSU issue, I decided to unplug all non-essentails from my PC. The optical drives, my extra hard drive, and the fans are now all unplugged.

The PC booted fine as expected. Upon getting into windows. I decided to run AtiTool to attempt to stress the GPU a bit. Amazingly, it worked. I was able to run ATITool without the computer freezing or resetting. I then tried FurMark and again it worked.

I went to push my GPU a little farther and use FurMark at my monitors resolution of 1680x1050 in fullscreen, and the system crashed.

Am I wrong to think this is a PSU problem? First I was unable to run any type of GPU test program. After unplugging the non-essentials, I was able to run AtiTool and Furmark as long as they were in windowed mode or a smaller resolution. But when I attempt to run them fullscreen, I get a system crash again. Perhaps I don't have quite enough power to test at full screen, thus it crashes.

Since these are GPU test programs, I am going to let ATITool run for a while to record my GPUs temps and to see if any issues arise with my GPU itself.

Also, while ATITool is running, I can here a slightly high pitched squeaking or scratchy noise coming from the back end of my PSU. Possibly another indication my PSU is going out.

Update 2:

ATITool has been running for 12min and counting with no errors, no freezing, and no resetting. I no longer think it is a problem with my GPU and I have a case of a PSU going out. My GPU is currently running at about 87-89C while ATITool is running. This seems hot but I am unsure if this is normal for a 8800GTS 640MB XXX. Does anyone know the normal temps for my GPU under 100% load?

Keep in mind no fans are currently on in the system. I have them unplugged still and the side cover is off. So temps may be a little hotter than they might normally be.

Update 3:

Just plugged everything back in as it normally would be, started up windows, immediately tried to run ATITool and instant crash.

Sorry for talking to myself here. I guess it helped me fix my own problem. If anyone still knows about an operating temp in the 80s for my GPU being okay or not I would appreciate it.
 

DrivingDegree

Distinguished
Dec 20, 2009
16
0
18,510
Well, like you said, the temp. of the GPU is 87-89, like i know i had the same problem but with a NVIDIA 8500GT (overheating) that's cuz i hat NVIDIA GPU's they're too hot, but, ya, is an overheating problem with your GPU because the max. temp. Is 86°C for NVIDIA, so i'll recommend you buy an ATi graphics card (ATi HD 4670 doesn't need extra power connector) and ATi is MORE FRESH THAN NVIDIA, my current temp. is 40°C with overclock and a resolution of 1440x900 on windows 7, so i guess that's fresh, and without OC, it's on 28-35°C so.... like i've saw playing GTA IV on high resolution and medium settings (95% load) the temp. is on 60°C.

I guess i helped you a lot and remember: The problem is in your GPU, not the PSU, max temps for NVIDIA is 86°C, and too you can download PCWizard 2010, is a very good program and you can monitorize ur temps of the CPU, GPU, Voltage from CPU, core temps., core loads, so it is a very good program.

^.--