IS&T RCS Tutorial - Introduction to OpenACC

OpenACC (for Open Accelerators) is a programming standard for parallel computing on GPU systems. It is designed to simplify GPU programming. The basic approach is to insert special comments (directives) into the code that offload computation onto the GPUs and parallelize the code at the levels of GPU blocks and threads. It is possible for programmers to create an efficient parallel GPU code with only minor modifications to a serial CPU code. This tutorial covers the major features of OpenACC through discussion, examples and hands-on exercises in C and Fortran. Some experience in C or Fortran programming is assumed. Understanding of parallel programming in general is helpful, but not required.