I have a software implementation in C that uses CUDA to speed up computations. Would like to have a go at OpenCL Can one use OpenCL with GCC? What can be done for Multicore and GPU Programming to speed up computations using parallel implementations with GCC?