How to configure OpenMP in the Intel IPP library to maximize multi-threaded performance of the Intel IPP primitives.
This article explains how to create your own OpenMP* thread mapping to OS proc IDs with Intel® C/C++ and Fortran compiler, by setting environment variable KMP_AFFINITY. It also provides some thread affinity demo results on when use different settings.
Recommended Settings for Calling Intel® Math Kernel Library Routines from Multi-Threaded ApplicationsRecommended settings for calling Intel MKL routines from multi-threaded1 applications
The Intel C++ and Fortran compilers for Windows* and Linux* provide 'legacy' and 'compatibility' implementations of the OpenMP THREADPRIVATE directive. The 'compatibility' option should not be used when everything is compiled by Intel compilers.
The newest releases of the Intel® C++ and Fortran Compilers support new features in the OpenMP* 3.1 Specification
An application compiled with -Qopenmp (Windows*)/-openmp (Linux*) and run on a host not containing a genuine Intel® processor may generate bogus OMP: Warning #71 and Warning #72.
Due to an issue with the default libthread library on Pardus Linux* version 2011.2, 64-bit, the debugger cannot detect which threading library is being used by the debuggee.
The compiler does not parallelize OpenMP loops that contain a "continue" statement in a C++ catch block inside the parallel region.
Sparse BLAS routines can be useful to implement iterative methods for solving large sparse systems of equations or eigenvalue problems
This article discussions parallelization and provides links that will help you understand your programming environment and evaluate the suitability of your app.