Affiliation:
1. Universidad Nacional de C´ordoba Facultad de Cs. Exactas, F´ısicas y Naturales - Lab. de Computaci´on Av Velez Sarsfield 1611, 5016 C´ordoba ARGENTINA
Abstract
Current supercomputers are composed by nodes containing a combination of general purpose computing units (CPUs) and specific mathematical coprocessors. In this way, GPGPUs or Xeon Phi cards are attached to the nodes to improve its performance. Both types of processors, CPUs and coprocessors, have many differences, like their architecture, the clock rate of the processors and the operation of the related memory. These are the main factors that conform an heterogeneous multiprocessor. A parallel program that wants to achieve the sum of the performance of both types of processors, must consider not only the complexity of the parallel algorithm, but also the differences in the architecture of processors, increasing its complexity. As a contribution on this problem, this paper presents a model of parallel execution based on Petri Nets, called PN-PEM, that is used not only to model a parallel algorithm, but also to execute it directly on a computer with heterogeneous multiprocessors. An asynchronous execution of tasks and a dynamic scheduler are the main characteristics that allow execute a parallel program on this type of parallel computer. Tests done on a multicore computer with two Xeon Phi cards reach the aggregate performance of both type of processors, confirming the quality of the model used.
Publisher
World Scientific and Engineering Academy and Society (WSEAS)
Reference18 articles.
1. Cedric Augonnet, Samuel Thibault, and Raymond Namyst. Starpu: a runtime system for scheduling tasks over accelerator-based multicore machines. Technical Report 7240, INRIA, March 2010.
2. SJ Bachega and DM Tavares. Applications of petri nets in distributed processing: a scoping study. In Proceedings of the International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA), page 109. The Steering Committee of WorldComp, 2015.
3. Alfredo Buttari, Julien Langou, Jakub Kurzak, and J. Dongarra. A class of parallel tiled linear algebra algorithms for multicore architectures. Technical Report 191, LAPACK Working Note, September 2007.
4. Michel Diaz. Petri Nets: Fundamental Models, Verification and Applications. ISTE Ltd - John Wiley & Sons, Inc., London, Hoboken, 2009.
5. Jack Dongarra, Mark Gates, Azzam Haidar, Yulu Jia, Khairul Kabir, Piotr Luszczek, and Stanimire Tomov. Hpc programming on intel many-integratedcore hardware with magma port to xeon phi. Scientific Programming, 2015, 2015.