Energy-aware parallelization flow and toolset for C code

Multicore architectures are increasingly used in emhedded systems to achieve higher throughput with lower energy consumption. This trend accentuates the need to convert existing sequential code to effectively exploit the resonrees of these architectures. We present a parallelizatiou flow and toolset for legacy C code that includes a performance estiniation tool, a parallelization tool, and a streaming-oriented parallelization framework. These are part of the work-in- progress EU FPT PHARAON project that aims to develop a complete set of techniques aud tools to guide and assist software development for heterogeneous parallel architectures. 'We demoustrate the effectiveness of the use of the toolset in au experiment where we measure the parallelization quality aud tune for inexperienced nsers, and the parallelizatinn flow and performance results for the parallelization of a practical example of a stereo vision appheation