Something wrong with my GPU

[aristotlezero1]

Hello I'm new here in this site, also advance sorry for my bad English

anyway here's my problem:

I bought an GTX 660 non-Ti a month ago
every time I play for about 20-30 mins, my games crash, I don't see any artifacts in the game it's just crashing
and after some crash, I can't launch google chrome, I got an error of APPCRASH
I also notice that sometimes, when the GPU hit around 60c and above, the crashing occurs
also I have a Rocketdock running in my desktop, whenever I browse rocketdock, I have a strange glitches in the icons

here's what I'm talking about rocketdock
http://i1063.photobucket.com/albums/t506/Yvonne_Lawson/PA300608.jpg
http://i1063.photobucket.com/albums/t506/Yvonne_Lawson/PA300602.jpg

as you can see in the picture, the text in the icons is not right. this happens when the game im playing crash

My first suspect is the GPU itself, second is my PSU but I believe a GTX 660 non Ti is not power hungry

My system specs
AMD Phenom II X4 965(stock clocks)
Asrock 880gm-LE motherboard
Asus GTX 660 non Ti
4gb ram
Aerocool Strike-X 600w PSU

Hopes someone help me ty.

 

[mocchan]

Which drivers are you using? If you're using the latest drivers, revert back to one where you didn't have any issues.

Secondly, I don't think your PSU would be the cause of this issue. However, with generic no-name PSU's you can't trust them to even provide the rated wattage. I would highly recommend swapping out out ASAP.

 

[aristotlezero1]

Thanks for your reply

Im using the latest drivers which is 310.33

I also reverted back to 306.93 and still no luck

 

[JJ1217]

Aerocool aren't exactly no-name, they make fantastic cases (X-Predator comes to mind, however I agree they aren't the vendor you would go to for Power Supplies. Although I also doubt its the issue.

 

[techgeek]

I would try something like OCCT GPU test, 3D Vantage, FurMark or Unigine. Really anything that will warm the card up and see if it crashes. OCCT GPU test or Furmark would be great as they run windowed by default. Then you can keep an eye on any monitoring software as the card warms up. Also with either of these two, it should warm the GPU up really quickly. If heat/GPU quality is really the issue, it should crash very quickly with either of these two tests. Even though 60 degrees isn't that warm especially for nVidia GPU's, I've seen cards start artifact and crash as similar temps.

My guess is the card is failing.