Already since 1998, the HPC team of the IT Center is engaged in shared memory parallelization with OpenMP. Since then, OpenMP is used in production on the large SMP computers of the IT Center with currently up to 128 cores. (The ScaleMP - service has been discontinued on July 31, 2014. It allowed OpenMP programming with up to 512 cores on a Cluster equipped with the vSMP software.) Recently, with the advent of multi- and manycore processors in the mass market OpenMP is gaining more and more steam. Also derivates of OpenMP (like Open ACC) are employed on accelerators like GPUs from NVIDIA and the upcoming Xeon Phi from Intel.
- You can find a short description of our OpenMP activities in the “early” years 1999-2006 here (pdf, in German).
- You can find some of our publications listed here (html) or here (pdf).
You can use the following OpenMP compilers on our machinery:
OpenMP Compilers at the IT Center
- Intel C++ and Fortran Compilers on Linux and Windows
- Visual Studio for C++ on Windows
- Oracle C++ and Fortran Compilers on Linux
- PGI Compilers for C, C++ and Fortran on Linux. The PGI-Accelerator Compiler also supports OpenACC (partially), CUDA-Fortran and proprietary PGI-Accelerator directives for NVIDIA GPUs
- GNU GCC for C, C++ and Fortran on Linux
Using OpenMP at the IT Center
- R. van der Pas, B. Chapman, G. Jost: “Using OpenMP”, MIT Press. ISBN-10: 0-262-53302-2 / ISBN-13: 978-0-262-53302-7.
- R. Chandra, L. Dagum, D. Kohr, D. Maydan, J. McDonald, R.Menon: “Parallel Programming in OpenMP”, Morgan Kaufmann Publishers, 2001
- Version 4.5 (November 2015)
- Version 4.0 (August 2013)
- Version 3.1 (July 2011)
- Version 3.0 (May 2008)
- Version 2.5 (March 2005)
- C/C++ version 2.0 (March 2002)
- FORTRAN version 2.0 (November 2000)
- C/C++ version 1.0 (October 1998)
- FORTRAN version 1.1 (November 1999 - incorporates April 1999 Interpretations and Errata)
- FORTRAN version 1.0 (October 1997)