Wayne D's Blog

Recently I ran into a problem with the screen going black on my EL6 Linux system under normal operating conditions. A search of /var/log/messages turned up the following message that occurred whenever the problem occurred.

messages-20130127:Jan 26 12:04:44 localhost kernel: NVRM: GPU at 0000:08:00.0 has fallen off the bus.

After trying a number of solutions suggested by others on the internet without success I used the nvidia-smi tool to determine that at idle the video card temperature was 65C when normally it should have been 35C. Under load the temperature would rise to 80C and above. It turns out the fan on the Nvidia GT430 cards we were using was of poor quality and the fan was barely turning on the video card. When the video card was replaced with another Nvidia card with a better quality fan the problem disappear.

If overheating isn't the cause of your problem check out some of the solutions at http://www.cyberciti.biz/faq/debian-ubuntu-rhel-fedora-linux-nvidia-nvrm-gpu-fallen-off-bus/