I’m getting the following error when I run deviceQuery over a remote desktop connection:
cudaGetDeviceCount FAILED CUDA Driver and Runtime version may be mismatched.
I’m running Windows 7 x64 with Tesla C2050 and the driver version is 263.06. I’m seeing this error when I have only C2050 installed as the primary display card. When I add in a separate video card, it works fine.
So, is it required to have a separate video card as primary card for CUDA applications to work over RDP/Session 0?