Introduction

PSkel is a high-performance framework for parallel skeletons. Using a high-level abstraction for parallel skeletons, PSkel releases the programmer from the responsibility of writing boiler-plate code for parallel programming in heterogeneous architectures, e.g., explicit synchronization and data exchanges between GPU memory and main memory. Furthermore, the framework translates the abstractions described using its application programming interface (API) into lowlevel C++ code compatible with Intel TBB, OpenMP and NVIDIA CUDA. PSkel's API is mainly based on a C++ template library that implements parallel skeletons and provides useful constructs for developing parallel applications. The framework provides an API for manipulating input and output data; specifying stencil computations; encapsulating memory management, computations, and runtime details.