OpenCL (Open Computing Language) is an open standard for leveraging the computing power of GPUs and multi-core CPUs. OpenCL makes it possible for you to write cross-platform code that can execute on a variety of CPUs and GPUs without having to use the language of a specific vendor or to map general-purpose code to a 3D graphics API such as OpenGL or DirectX. OpenCL provides an abstraction layer that allows your general-purpose code to run in parallel on all the GPU and CPU cores on a system. The OpenCL technology is specifically designed for use with applications that need extra computing power, that must be portable, and that need to respond readily to evolving improvements in hardware. OpenCL can benefit scientific applications, image processing, signal processing, and other programs that require large amounts of parallel processing.
This guid explains how to use OpenCL to exploit the computing power that resides in the many processing cores in a Macintosh computer’s graphics processing unit (GPU) and central processing unit (CPU).