Graphics Card Troubleshooting Software?

[Macheez]

Well, I'm actually posting a problem that I have and am hoping someone has a shred of hope of a solution.

I purchased this card in November 2010:
EVGA 01G-P3-1452-TR GeForce GTS 450 SuperClocked Video Card - 1024MB GDDR5, PCI-Express 2.0, Dual DVI, Mini HDMI, SLI, DirectX 11, Fermi +99MHz Overclock to 882MHz

Problem:
Major artifacts occur only when playing intense games.

Troubleshooting already tried:
-Verified using latest drivers.
-Reinstalled operating system and graphics drivers.
-Unseated card and checked for scorching on teeth.
-Downloaded EVGA OC scanner which goes nuts roughly 10-15 seconds into the test.

I don't want to spend a bunch of money on any software but if there is something very reliable in troubleshooting the cause(and solution) then I am quite interested.
I'm still able to use this card without problems when not playing an intense game.

My suspicion is that this card has a hardware failure.  The warranty expired at the end of November 2012.
Thanks in advance for any recommendations/advice/opinions!
Mac-n-Cheez
 

[Dusaro]

To me it sounds like a power or heat issue.

After how long do these major artifacts occur?

What is the max temperature the card is reaching while these artifacts are appearing?

Are the Fans/Coolers working?

Are all the cables from the PSU connected correctly? Have any come loose?

[Macheez]

Thanks Dusaro,

I agree that this is either a heat or power problem.  And I'm glad you said power because I was just focusing on heat.
I opened the machine again to double-check everything before answering.

After how long do these major artifacts occur?

Using the EVGA OC Scanner, the major artifacts begin after 40-45 seconds into the test.  Eventually, the test crashes from the issue.

What is the max temperature the card is reaching while these artifacts are appearing?

It starts at 64C and then increases to 81C before the test crash.  The fan speed goes from 41% and climbs with the temperature change.

Are the Fans/Coolers working?

All fans are working.  4-120mm fans, GPU fan, and the CPU fan.

Are all the cables from the PSU connected correctly? Have any come loose?

Just to be sure I unplugged/re-plugged the power connection to the graphics card.


I just cannot seem to pinpoint if this is a permanently damaged card or if I just need to tweak a setting.  Is it possible the GPU fan is not kicking in fast enough?
When I first installed this card 2 years ago it worked without any problems on the same games I play now and at the games highest graphic settings.

I'm really scratching my head on this one.
Mac-n-Cheez

[Totoro]

Have you added any additional cards / drives / etc in the pc lately? could it be your PSU is being strained?

-Nerd

[Macheez]

Good question Nerdstrom,

The answer is a simple no.  I have no other cards on the system and I have not added any other devices including a printer.

The PSU is a 450W which I have used since the beginning.
The only other source pulling power that I haven't mentioned is a DVD RW drive and the Hard drive.
Other than that I don't have any reason to believe the PSU is hurting for output.

Do you think I should recalculate the usage?
Mac-n-Cheez

[Totoro]

Well if it used to work fine, and you haven't added anything new to the computer, it's probably not that... Hmm.

[Dusaro]

Is it possible to screenshot these artifacts? e.g with print screen or a camera.

Try set up a desk fan to blow into the computer or onto the card.

Try under-clocking the card and see if the artifacts still occur.

[BFM_Kiwi]

FWIW

Maximum GPU Temperature  100 C
Maximum Graphics Card Power 106 W
Minimum System Power Requirement 400W

http://www.geforce.com/hardware/desktop-gpus/geforce-gts-450/specifications

[Macheez]

Thank you everyone for your suggestions!  The card is working again!

Short story:
New thermal paste & the correct 6-pin power cable has it working like right out-of-the-box!

Long story:
@Nerdstrøm & @Kiwi
You two got me really thinking maybe the hyperlink:Corsair 450W PSU was just barely enough power for everything.
I even re-checked using a hyperlink:PSU calculator from Asus which is known to assign more power consumption than needed.
The calculator was saying 450W should be enough but I was starting to doubt and think I should upgrade to a 500-600W just to be safe.
With the other research and little tidbits I gleaned there was a "smart" moment when I unplugged the cable from the graphics card.  I looked at the end of the cable and realized it only had 5-pins in a 6-pin head while the card showed it was expecting 6-pins.  And guess what?  I happened to look at my power cables more closely and found a 6-pin that even said PCI-E on it.  Did I ever have a "DUH!" moment... LOL

@Dusaro & @ConVicT 626
Your thoughts on the cooling were helpful because I found EVGA's software on the original disc.  It's called Precision and I upgraded to their current version, Precision X.
I had a screenshot ready to post for you Dusaro.  But after seeing the Stress Tests continually go bonkers at 60ºC and the heat just crank up until the system throws an error, well, heat definitely seemed an issue.
At first I played with the Precision software to adjust the fan to run higher and much sooner than its defaults.  But that still wasn't helping much.  I finally decided it couldn't hurt to open up the GPU.  When I saw the manufacture's gray paste, it looked great and fully covered.  But I still wiped it all off and "liberally" applied my own that I purchase from Radio Shack.  So after all that I ran the stress test and it ran so beautifully on the cards default overclocked settings.  Not even one article showing out-of-whack.

So a BIG THANKS to all four of you!  This experience has taught me and improved my troubleshooting skills.
AND!!!  I can return the $69.99 EVGA GeForce GT 640, PCIe 3.0, 2Gb DDR3

See you on the servers!
Mac-n-Cheez

[Dusaro]

Great to see you managed to fix it, and I must agree, always replace the thermal paste, probably the first thing I learnt when working with pre-built computers and components.