Alienware m17 GPU overheating "-1% fan" running cudahashcat. Details in post.

[Sleet]

This is the readout before the 90c shutdown.

HWMon.GPU.#1...: -1% Util, 89c Temp, -1% Fan
HWMon.GPU.#2...: -1% Util, 77c Temp, 100% Fan

The GPU #2 never gets above 80c and I can run it for hours if I shut off GPU #1.

I have tested the fans in the BIOS diagnostics program and everything checks out. I have noticed that when the GPU fans kick on in windows it sounds like the damn thing is gonna take off (not in a bad way) but in The-Distribution-Which-Does-Not-Handle-OpenCL-Well (Kali) it's not near as pronounced.
I get about 4 minutes from start to program shutdown.
I have a CORE i7 processor and 2 Geforce GTX 675M GPUs.

I have googled this to death and have only found one guy on Twitter with the same readout joking about cooking an egg on his GPU, which was worse than finding nothing at all.
I have the newest proprietary drivers installed and am running The-Distribution-Which-Does-Not-Handle-OpenCL-Well (Kali) 1.0.7

I also tried this on the windows program with the same results.

[Mangix]

I don't think it's a good idea to run hashcat on laptops with powerful GPUs. In any case, 90C for a GPU is way too much.

[Sleet]

I don't think it's a good idea to run hashcat on laptops with powerful GPUs. In any case, 90C for a GPU is way too much.

Thank you, that is where they shutdown. I agree it seems like a little much. 80c is as low as I can keep the gpu with the working fan, do you think that is acceptable?

[Mangix]

Absolutely not. The hottest GPU that I've ever owned never went above 70C when using hashcat. Keep in mind that heat as well as voltage are what lowers the life of a GPU.

Just use a desktop. Or wait for maxwell to come to laptops. Or if you really want to use your current laptop, you can try lower -n values.