Hello everyone. I have been trying to jump into cuda over the past couple of days with unfortunately no luck. My platform is a laptop with a 520M video card in it. (I’m not expecting a whole lot of performance I just want to see it work). I’m running gentoo linux and have the cuda developer package installed. I have followed some tutorials on both the nvidia development website as external tutorials and I seem to be running in to the same problem. My programs compile and run but it doesn’t seem to properly communicate with the card. It just returns a zero no matter what the calculation.
I downloaded sample code and have the same problem.
I even wrote a program which doesn’t execute any kernels, but simpily sends data to the card and attempts to retrieve it. No luck.
I can’t understand what would be causing this. The only thing I can think of, and I’m going to feel pretty dumb if this is the case. Does the card have to be dedicated? Does the fact that it supports my graphics prevent it from doing any cuda calculations? If thats not it, any other suggestions?