Geodesic
Agent-Oriented Approach to Simulate Exaflop Supercomputer with Application to Distributed Stochastic Simulation
Vestnik Ûžno-Uralʹskogo gosudarstvennogo universiteta. Seriâ, Matematičeskoe modelirovanie i programmirovanie, no. 12 (2012), pp. 93-106 Cet article a éte moissonné depuis la source Math-Net.Ru

Voir la notice de l'article

A possibility of using an agent-oriented simulation system for solving a variety of problems that arise in design and implementation of exaflop supercomputers consisting of ten and hundred millions of computational nodes is discussed in the paper. We suggest two-lewel decentralized scheme of computations control and the corresponding simulation model in which all the computational nodes are distributed over computational domains controlled by their control agents. Master control agent distributes a flow of big problems over computational domains and manages common resources. Monte-Carlo method considered to be promising to use on exaflop supercomputers is given as an example of highly scalable algorithm. In this method, is essential that the lager sample size of independent realizations, the higher accuracy of estimating. We also suggest a parallel pseudorandom numbers generator suitable for large-scale computations with Monte Carlo method. When distributing stochastic computations over different nodes it is possible to simulate different sample volumes on different nodes using statistically optimal technique of results averaging. Naturally, an amount of computer resources available on each node must be quite enough to simulate the realizations effectively. The described algorithm of distributed stochastic simulation is asynchronous one and can be scaled to a practically infinite number of nodes using the described parallel pseudorandom numbers generator. An example of the highly scalable application utilizing distributed stochastic simulation on up-to-date teraflop supercomputers is the program library PARMONC. Also, the multi-agent simulation is used for the prediction and processing of possible failures of computational nodes. Architecture of dynamic system of failures prediction is given. The system consists of the agent for different purposes; each agent is playing its role to achieve the common goal.
Keywords: agent-oriented simulation, exaflop supercomputer, Monte Carlo method, distributed stochastic simulation, parallel computations. 
@article{VYURU_2012_12_a8,
     author = {B. M. Glinsky and A. S. Rodionov and M. A. Marchenko and D. I. Podkorytov and D. V. Weins},
     title = {Agent-Oriented {Approach} to {Simulate} {Exaflop} {Supercomputer} with {Application} to {Distributed} {Stochastic} {Simulation}},
     journal = {Vestnik \^U\v{z}no-Uralʹskogo gosudarstvennogo universiteta. Seri\^a, Matemati\v{c}eskoe modelirovanie i programmirovanie},
     pages = {93--106},
     year = {2012},
     number = {12},
     language = {ru},
     url = {http://geodesic.mathdoc.fr/item/VYURU_2012_12_a8/}
}
TY  - JOUR
AU  - B. M. Glinsky
AU  - A. S. Rodionov
AU  - M. A. Marchenko
AU  - D. I. Podkorytov
AU  - D. V. Weins
TI  - Agent-Oriented Approach to Simulate Exaflop Supercomputer with Application to Distributed Stochastic Simulation
JO  - Vestnik Ûžno-Uralʹskogo gosudarstvennogo universiteta. Seriâ, Matematičeskoe modelirovanie i programmirovanie
PY  - 2012
SP  - 93
EP  - 106
IS  - 12
UR  - http://geodesic.mathdoc.fr/item/VYURU_2012_12_a8/
LA  - ru
ID  - VYURU_2012_12_a8
ER  - 
%0 Journal Article
%A B. M. Glinsky
%A A. S. Rodionov
%A M. A. Marchenko
%A D. I. Podkorytov
%A D. V. Weins
%T Agent-Oriented Approach to Simulate Exaflop Supercomputer with Application to Distributed Stochastic Simulation
%J Vestnik Ûžno-Uralʹskogo gosudarstvennogo universiteta. Seriâ, Matematičeskoe modelirovanie i programmirovanie
%D 2012
%P 93-106
%N 12
%U http://geodesic.mathdoc.fr/item/VYURU_2012_12_a8/
%G ru
%F VYURU_2012_12_a8
B. M. Glinsky; A. S. Rodionov; M. A. Marchenko; D. I. Podkorytov; D. V. Weins. Agent-Oriented Approach to Simulate Exaflop Supercomputer with Application to Distributed Stochastic Simulation. Vestnik Ûžno-Uralʹskogo gosudarstvennogo universiteta. Seriâ, Matematičeskoe modelirovanie i programmirovanie, no. 12 (2012), pp. 93-106. http://geodesic.mathdoc.fr/item/VYURU_2012_12_a8/

[1] R. O'Keefe, “Simulation and expert systems — a taxonomy and some examples”, Simulation, 46:1 (1986), 10–15 | DOI

[2] Rodionov A. S., “Intellectual Simulation — New Approach in Imitation Systems (Review of Recent Publications)”, Expert Systems and Data Bases, Novosibirsk, 1988, 19–35

[3] Wooldridge M., Introduction to MultiAgent Systems, John Wiley Sons Ltd., England, 2002

[4] T. Oren, L. Yilmaz, “On the Synergy of Simulation and Agents: An Innovation Paradigm Perspective”, International J. of Intelligent Control and Systems, 14:1 (2009), 4–19 | MR

[5] R. L. Bargodia, K. M. Chandy, J. Misra, “A Message-based approach to discrete-event simulation”, IEEE Trans. on Soft. Eng., 13:6 (1987), 654–665

[6] Karpov Yu. G., Imitation Modeling of Systems. Introduction to Simulation with Any-Logic 5, BHV-Petersburg, SPB., 2005

[7] E. Niewiadomska-Szynkiewicz, A. Sikora, “Algorithms for Distributed Simulation–Comparative Study”, PARELEC 2002 (Warsaw, Poland), 261–266

[8] Marchenko M. A., Mikhailov G. A., “Distributed Computing by the Monte Carlo Method”, Automation and Remote Control, 68:5 (2007), 888–-900 | DOI | MR | MR | Zbl

[9] Mikhailov G. A., Voytishek A. V., Numerical Stochastic Simulation. Monte Carlo Method, Publishing Center «Akademia», 2006

[10] M. Marchenko, “PARMONC — A Software Library for Massively Parallel Stochastic Simulation”, PACT 2011, LNCS, 6873, 2011, 302–315

[11] Gordienko D. V., “Developing and Modeling the System of Controlling the Power Consumption of Cluster Computing Systems”, High performance computations on cluster systems, Proceedings of International Conference (Vladimir, 2009), 126–130