CUDA in root?

Hi

I want to run my CUDA code from root’s cint , what is the best way to accomplish this ? I have read about the findCUDA.cmake script , but I would prefer if it could be done without having to rebuild .
I think loading the CUDA shared libraries with gSystem->Load is to be done, followed by some sort of interface , yes ? Please suggest some way to implement this. (or maybe something else)

Thanks

Hi,

Can you be more specific in the kind of code you are trying to interpret and how it currently failed.

Cheers,
Philippe.