I am having a long saga in my efforts to make my GPU program in parallel. I do not use my GPU for graphics of gaming, I bought it for parallel programming. But never thought it would be such a difficult task.
My system is:
Windows 7, 32 bit
Quadro FX 3700
Cuda Toolkit, Computing SDK 4.0
Visual Studio 2008
Exmaples ran but VS didn’t work for new projects!
Visual Studio 2010 Express
Visual Studio 2010 Ultimate
Thought it might be VS’ non-compatibility issue, so tried:
Then, I changed SDK to 4.1. I can execute kernels…finally. But, the simple vector multiplication code returns the same vector values that I insert in the function.
I tried C and then C++.
What can I do? Is there any platform which works with CUDA on my system? Please, save my life. :)