I’m ver new in this area (GPU parallel programming), and I have a fundamental question. I’ working in an application which is written in multi-language C/C++ & Fortran using visual studio. I downloaded the CUDA toolkits which include CUDA C and OpenCL. Until now I don’t now what are the major differences between CUDA C and OpenCL. At the moment, I need to take a decision which one to use. Can any one tell me what are the major advantages of one of them on another.
Another question, the toolkits include programing guides and samples for each of CUDA C and OpenCL but I didn’t find any documents explaining how to build the solution in visual studio. I can copy *.cu file from the samples but I don’t know how to build my solution. I’m wondering if anyone can give me a link to documents which explain this issue.
Can any one tell me what are the major advantages of one of them on another.
CUDA is only runs on NVidia hardware. OpenCL is supported by AMD, Apple, IBM, Intel, NVidia, Qualcomm and many others. Other than that, they have similar functionality.
Obviously I am biased.