[Nvidia 331.x - 334.21] GT 635M - straight overheating

my specs:

Arch Linux 64-bit
kernel: 3.13.5-1-ARCH
xorg: 1.15.0-5
xrandr: 1.4.1-1

using primus: 2013 12 26

laptop model: Asus K55VJ-SX069H
graphics: NVIDIA® GeForce with CUDA® 635M 2GB DDR3 VRAM
CPU: INTEL Core i5 Ivy Bridge (3210M)

my problem

When I start any application via nvidia card card constantly overheats until it shutdown whole laptop because of too hot GPU. It definitely seems like a driver-related issue since previous versions worked just fine (sadly I can’t use them anymore because of new xorg version). It looks like driver just ignores graphic card’s fan (and don’t even use it) according to this output of nvidia-smi and differences between some tests on Windows 8 and Linux. I hope you’ll fix it as soon as possible since I have this problem already few months and I can’t play some games since integrated intel solution is too slow.

Wed Mar  5 17:28:03 2014       
| NVIDIA-SMI 334.21     Driver Version: 334.21         |                       
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|   0  GeForce GT 635M     Off  | 0000:01:00.0     N/A |                  N/A |
| N/A   82C  N/A     N/A /  N/A |    280MiB /  2047MiB |     N/A      Default |
| Compute processes:                                               GPU Memory |
|  GPU       PID  Process name                                     Usage      |
|    0            Not Supported                                               |

Also sorry for my funny english, i’ts not my native language.

nvidia-bug-report.log.gz (70.5 KB)

Added GRUB_CMDLINE_LINUX=“acpi_osi=Linux” to my grub. Fan working, but it still overheats until it shutdown. After this improvement I did few test and it looks like nvidia card overheats other components around and CPU overheat cause reboot. I am sure that it’s not CPU fail since previous drivers worked just fine.

more specs:

LVDS (internal): 1366x768 @ 60Hz
HDMI (external): 1920x1080 @ 60Hz

around 4°C difference between card and other parts

Can you reproduce this problem in Windows 7/8.1 with latest NVIDIA drivers?

I smell a hardware problem here, like e.g. the temperature sensor doesn’t work properly.

Good point, but sadly under W8 with latest driver is everything ok.

today I cleaned up whole laptop and also replaced thermal paste on both CPU and GPU (nvidia),
sadly it still do the same thing and temperature sensor must work since nvidia-smi show temperature
Devs, please fix that, I’d like to use my nvidia card on linux too.

Does the new 337 driver fix the issue by any chance? There’s no fan control option under nvidia-settings?