GPU Throttling during Furmark Burn-In on my ASUS G75VW

I'm not a gamer, so would usually not realize if there is a problem with my GPU. But I was running the Furmark 15 minute Burn-In test this morning, and it seems my GPU is severely throttling. It does this from the second I start the test until the test ends (see below).

Furmark v1.10.6 - BURN-IN benchmark, 1920x1080 (0X MSAA)
Frames: - time: - FPS:12 (min:11, max:13, avg:12)
>OpenGL renderer: GeForce GTX 660M/PCIe/SSE2
>GPU 1 (NVIDIA GeForce GTX 660M) - core: 950MHz (THROTTLING: -545MHz), mem: 2499MHz, temp: 74C, GPU load: 99%, mem load: 12%, fan: 0%

-----------------------
I have everything set to high performance in Power4Gear, and maximums set to 100%. Another strange thing: if you'll notice, it shows my fan at 0%. But when running the test, I can hear and feel the fan spinning pretty fast.

Anybody have a clue as to why these things are happening, or possibly have a cure?
 

ingtar33

Honorable
Dec 17, 2012
249
0
10,910
that's what nvidia drivers do on purpose. Nvidia intentionally throttles it's gpu when it detects furmark, in order to prevent heat related damage to the gpu.

Your gpu temps are fine because the driver is working properly. i think they started doing that with the 400 series gpus...
 
Many thanks for the reply!! It was very informative.

One more question:

As I'm not a gamer, is there some simple task I could run to see if my GPU ever moves off of the apparently base clock of 405MHz while idling? I'm not looking to benchmark anything, or stress the GPU to 100%. Just looking to make sure it isn't defective. I can use CPU-Z and Speccy to monitor the clock speed, just need something simple to push the GPU a little.
 

ingtar33

Honorable
Dec 17, 2012
249
0
10,910


well... just play a game. typically mobile gpus will jump off their idle speed the moment they're stressed at all.