RE: FATAL ERROR: CUDA error in cudaGetDeviceCount on Pe 0 (thomasASUS): CUDA driver version is insufficient for CUDA runtime version

From: Roberts, Jason (
Date: Sat Sep 22 2012 - 20:34:50 CDT

Hi Thomas,

I we are running 2 laptops using Ubuntu, one with a 540M and the other with a 640M. To get around the driver issues associated with the "Optimus" technology I had to go into the BIOS at start-up and change the graphics card option to "Discrete" . This means that the system no longer uses the on-board Intel HD graphics chip and runs entirely on the nVidia GPU, which solved our problems. From my understanding nVidia are in the process of deriving a solution for using Optimus in a Linux environment, but until then the only options I know of are independent solutions which I've found cause problems when running some programs which use CUDA extensions.

So far I have found no real problems with running the systems on the nVidia GPUs permanently, the systems are more stable but there is a minor drop in battery time.



Jason A. Roberts
Senior Medical Scientist
National Enterovirus Reference Laboratory
WHO Poliomyelitis Regional Reference Laboratory
VIDRL, 10 Wreckyn Street,
North Melbourne, Australia, 3051
Phone: +613 9342 2607
Fax: +613 9342 2665
email: (lab enquiries)
web site:

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:22:06 CST