This might sound a little bit silly, but I haven’t found a definitive answer. I’ve got an application that would largely benefit from the scalability of OpenCL. But how would I go about with the existing code I have, to implement OpenCL? Would I have to use a wrapper? Do I have to port the code over to OpenCL?
Share
A wrapper is not sufficient. You will have to rewrite your code in OpenCL C and most likely redesign your algorithms for parallel processing (running on multiple cores/threads). There is no free lunch.