Posts

Showing posts from January, 2013

OpenMP an API to enable easier parallel programming in C, C++

OpenMP is an API to enable easier parallel programming in C, C++, and Fortran on shared-memory machines.
It’s widely supported on various system architectures and operating systems.
OpenMP is implemented primarily through compiler directives, which can be easily added to an existing serial C/C++ program.  These directives look like comments in your code.
The compiler can be told to ignore or activate these directives which in turn parallelises the code block with the statements.
This means that the conversion of your sequential C/C++ program to one that takes full advantage of a shared memory in a parallel computing system is a lot easier and can be taken in stages therefore you can test performance improvements step at a time.
To find out more http://openmp.org/wp/