CUDA code crashing or not finding NVS cards

I recently added an NVS 810 to a system that already has an NVS 310 in it. The system is running CentOS 7.
I used the .run installer to install CUDA and included drivers. That appeared to install without error.
However when I try to run the deviceQuery in the sample code I get a the following:

./deviceQuery Starting…

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

cudaGetDeviceCount returned 2
-> out of memory
Result = FAIL

I’m am uncertain where the problem is.