Scientific and Technical Journal


ISSN Print 2221-3937
ISSN Online 2221-3805

The coarse-grain parallel algorithm of modular neural networks training with dynamic mapping onto processors of parallel computer is described in this paper. Parallelization of the algorithm is done on parallel computer Origin 300 using MPI (Message Passing Interface) technology. Efficiency estimation of this parallel algorithm using MPE (MPI Parallel Environment) library has showed its disadvantage connected with calculation of a wall, not a processor computational time of parallel routine. The appropriate modification of the MPE library allows correctly define of a parallelization efficiency of a parallel routine.

  1. Trobec, R. Parallel Computing: Numerics, Applications, and Trends, Springer / R. Trobec, M. Vajteršic, P. Zinterhof. – 2009, 520 p. [in English].
  2. Chang, L.-C. An efficient parallel algorithm for LISSOM neural network, Parallel Computing / L.-C. Chang, F.-J. Chang. – 2002. – Vol. 28. – No. 11. – Р. 1611 – 1633 [in English].
  3. Estévez, P. A.A scalable parallel algorithm for training a hierarchical mixture of neural experts /P. A. Estévez, H.Paugam-Moisy, D.Puzenat et al. // Parallel Computing. – 2002. – Vol. 28. – No. 6. – P. 861 – 891 [in English].
  4. Vampir – a Tool for Performance Optimization [Electronic resource]. –Access mode:– 15.01.2013 [in English].
  5. Góes, L.F.W.Performance Analysis of Parallel Programs using Prober as a Single Aid Tool /L.F.W. Góes, L.E.S. Ramos, C.A.P.S.Martins //Proceedings of the 14th Symposium on Computer Architecture and High Performance Computing. – 2002. – P. 204 – 211 [in English].
  6. TAU Performance System® - a portable profiling and tracing toolkit for performance analysis of parallel programs written in Fortran, C, C++, Java, Python [Electronic resource].–Access mode: – 15.01.2013 [in English].
  7. Moore, S. Review of Performance Analysis Tools for MPI Parallel Programs / S.Moore, D. Cronk, K.S. London, J.Dongarra // Proceedings of the 8th European PVM/MPI Users' Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface. – 2001. – P. 241 – 248[in English].
  8. MPE– a software package for Message Passing Interface programmers[Electronic resource]. –Access mode: http://www.mcs.anl. gov/research/projects/perfvis/download/index.htm#MPE– 15.01.2013 [in English].
  9. MPICH - a high performance and widely portable implementation of the Message Passing Interface standard [Electronic resource].–Access mode:– 15.01.2013 [in English].
  10. Turchenko, V. Parallel Algorithm of Dynamic Mapping of Integrating Historical Data Neural Networks / V. Turchenko // Information Technologies and Systems. – 2004. – Vol. 7. – No. 1. – P. 45 – 52 [in English].
  11. Sachenko, A. Instrumentation for Data Gathering / A. Sachenko, V. Kochan, V. Turchenko // IEEE Instrumentation and Measurement Magazine. – 2003. – Vol. 6. – No. 3. – P. 34 – 40 [in English].
  12. A High Performance Message Passing Library [Electronic resource].–Access mode:–15.01.2013 [in English].
  13. Turchenko, V.Enchanced Method of Historical Data Integration Using Neural Networks / V. Turchenko, V. Kochan, A. Sachenko, Th. Laopoulos // Datchiki i Sistemy, Moscow, Russia. – 2002. – № 7 (38). – P. 35 – 38[in Russian].
  14. Performance Visualization for Parallel Programs [Electronic resource]. –Access mode:–15.01.2013[in English].
Last download:
25 Jan 2020

[ © KarelWintersky ] [ All articles ] [ All authors ]
[ © Odessa National Polytechnic University, 2014-2018. Any use of information from the site is possible only under the condition that the source link! ]