'cudaGetDeviceCount FAILED' with newest everything..

On some machines I simply can’t get the most vanilla example (deviceQuery) to work.

It’s an older 4.x CentOS that’s had everything maintained, with updated Kernel 2.6.18… The newest NVidia kernel modules build and load fine:
“NVRM: loading NVIDIA UNIX x86_64 Kernel Module 260.19.12 Fri Oct 8 11:17:08 PDT 2010”

Installing Cuda 3.1, the SDK examples build fine with no complaint. I set LD_LIBRARY_PATH to the cuda directory so it looks like everything’s going well, no complaints about LD stuff, but then upon running deviceQuery I get the output:

rosaleen : ./deviceQuery
./deviceQuery Starting…

CUDA Device Query (Runtime API) version (CUDART static linking)

cudaGetDeviceCount FAILED CUDA Driver and Runtime version may be mismatched.

FAILED

Press to Quit…

The RW permissions on /dev/nvidia0 and /dev/nvidiactl are fine - though failing at this point I think is before those even matter.

Any suggestions would be greatly appreciated, as this is keeping my company from looking further into Cuda…

Thanks,
Pat

Also I should add - I have searched this forum for previous responses to this question (there’s a bunch) - and I can’t find anything I haven’t tried or that is relevant.

Just to give an update for all those people who didn’t seem to care ;) In case anyone runs into this in the future, it’s an obvious answer…

I was attempting to test on an FX 550, which isn’t supported - as such even commands like getDeviceCount will fail (instead of returning 0 devices…).