[renegaderebel]

Hello everyone,
Last weekend my power supply failed on me while the PC was in use. After clearing the smoke and looking in the case, it didn't appear any of the other components were physically damaged. I made a quick trip to the local electronics store and grabbed a new supply unit, upgrading from a 500 watt to a 600 watt. After putting everything back together everything appeared normal. I could use typical apps(internet browsing, music etc.) However, I now get extremely low FPS when I'm using any graphics intensive programs(games etc) that worked flawlessly before.

I have a GeForce 8800 GT, Intel Core 2 2.13 Ghrz, and 4 GB of Ram. Since my only issues crop up when I'm running graphics programs, I can only assume the power supply damaged the GPU. I've 3dMark06 and scored thousands less then it predicted I should be.

Is there any way to validate the hardware of my card? And can anybody recommend me a cheap replacement? I'm not playing hot off the line games or anything.

Thanks in advance.

[Supertrucker]

If it loads up the games and visually it looks ok (no artifacts ect) but is simply much slower than it should be, i would start checking settings and software before i assumed the card was bad. Try going through your bios and redoing settings you had previously (if any) uninstall and reinstall your graphics drivers, check temps for both your cpu and video card, mabye a fan died/was taken out by the psu and your computer is temp throttling now. Failing all that i would do a clean instal of windows (or whichever OS you use) before i even thought about replacing the card.

[renegaderebel]

I've tried turning all the settings down. What made me thing it was hardware and not the game was I tried a different game(Portal) that used to work flawlessly on max settings and had similar slowdowns. However, I don't have any graphical artifacts or visual issues, so your presumption that it might not be the card has merit. I have watched the temps on the CPU/GPU while I'm running benchmarks and nothing ever goes above 90 C. The next thing I'm going to move the games to a new hard drive and attempt to run them from there. Also, is there anyway I can validate the new Power Supply is giving the correct wattage?

[CompuTamer]

90C is pretty hot for an 8800.

There's a good chance that the card could be throttling if it does hit those temps often. I don't think that happens till 100C though.

Only way to do that is to load test it with dedicated hardware. (basically a bunch of mosfets that aren't meant to do anything more than produce heat and eat power)

[renegaderebel]

Ok so after a through cleaning, double check my cables, and make sure the fans were working, I ran FurMark. Using Burn-in mode, I let it run for an hour. The hottest it ever got was 96 degrees with the majority of it being spent at the 89-90 mark. No visible artifacts popped up while running the test. It hung around 14 FPS for the entirety of the test.

One thing I noted was if I stopped the render of the furry circle, the FPS jumped up to 550 FPS and the temp dropped sharply. Even if I turned the post-FX option, with the furry render off it stayed at 500 FPS.

[JeremyCT]

Does GPU-Z show the card clocking up and down as it should?

[terran2k]

if your GPU was damaged, you'd see it on the screen. if it's running slow maybe a driver got corrupted when the power supply blew.
uninstall your video driver and re-install it.
out of curiosity, what power supply did you buy?

[renegaderebel]

The core clock never moves from 600. The memory load stays below 50% and doesn't rise. The fan speed does kick up as the chip gets hotter. I'm starting to wonder if it could be a GPU memory issue. When I start a game it will work fine for about 15-20 seconds and then start having periodic slowdowns.

I have uninstalled the drivers and reinstalled the latest with no change. I'm starting to wonder if its something besides the graphics card, but I don't experience any performance issues unless I'm running a game or GPU test. I ran memtest86+ through 200% coverage and it reported no issues. I ran prime95 for 8 hours today and it had no errors.

I'm at my wits end. I can't afford to go out and buy new parts, especially if I don't know which ones are causing the problem

PS. I got the coolmax v600 power supply.

[Ethan E]

You could try swapping cards with a friend to see if his card worked better in your system, and also if your card worked in his system. If your card is really broken, it will give the same bum results in his system.

I had a PSU go bad 4-5 years ago and it was a huge pain. It ended up killing my GPU, 2 sticks of ram, and my MOBO. Now I NEVER skimp on PSU. You dont need to go overboard on power, but never skimp on quality.

[King107s]

A long time back I had a PSU fry a video card (actually two video cards the same way) and this is the artifacting effect you will get....

[renegaderebel]

Yeah I don't have any artifacting at all. That like a crummy deal to me.

I don't know anybody else who has a PCI-E graphics card. What I'm thinking I might do is buy a new one and if it doesn't fix the issue just return the card.

[renegaderebel]

Well replacing the graphics card did not solve the problem. I knew that'd be too easy.

[SMOKEU]

I'm guessing it's a faulty motherboard.