Chapter 9. Program parallelization using OpenMP

The latest versions of the IBM C and C++ compiler products for AIX support OpenMP, which parallelizes your C and C++ applications. By inserting simple OpenMP directives^[1] (start with “#pragma omp”) into your source code, your application can be parallelized and efficiently performs on SMP systems.

^[1] In addition to the OpenMP directives, the IBM C and C++ compiler products for AIX also have been supporting IBM SMP directives to parallelize your C and C++ applications. See Appendix E, “Supported IBM SMP directives” on page 483 for more information about the IBM SMP directives.

This chapter explains how to use the OpenMP support of the IBM C and C++ compiler products for AIX.

The OpenMP support on AIX internally uses the POSIX threads explained in Chapter 8, “Introduction to POSIX threads” on page 275. Although the application programmer are not exposed to the POSIX threads when programming C and C++ applications with OpenMP, several concepts in the POSIX threads programming model are sometimes beneficial.

For further information about the OpenMP support provided by the compiler products, please refer to C for AIX Compiler Reference, SC09-4960 and VisualAge C++ for AIX Compiler Reference, SC09-4959.