Annals of Operations Research

, Volume 53, Issue 1, pp 249–285 | Cite as

Parallel simulation today

  • David Nicol
  • Richard Fujimoto
Article

Abstract

This paper surveys topics that presently define the state of the art in parallel simulation. Included in the tutorial are discussions on new protocols, mathematical performance analysis, time parallelism, hardware support for parallel simulation, load balancing algorithms, and dynamic memory management for optimistic snchronization.

Keywords

Parallel processing discrete-event simulation synchronization protocols memory management load balancing performance analysis 
This is a preview of subscription content, log in to check access.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    I.F. Akyildiz, L. Chen, S.R. Das, R.M. Fujimoto and R. Serfozo, Performance analysis of Time Warp with limited memory,Proc. 1992 ACM SIGMETRICS Conf. on Measuring and Modeling Computer Systems, Vol. 20 (1992).Google Scholar
  2. [2]
    H. Ammar and S. Deng, Time Warp simulation using time scale decomposition, in:Advances in Parallel and Distributed Simulation, Vol. 23, SCS Simulation Series (1991) pp. 11–24.Google Scholar
  3. [3]
    R. Ayani, A parallel simulation scheme based on the distance between objects,Proc. SCS Multiconf. on Distributed Simulation, Vol. 21 (1989) pp. 113–118.Google Scholar
  4. [4]
    R. Baccelli and M. Canales, Parallel simulation of stochastic Petri nets using recurrence equations, ACM TOMACS 3(1993)20–41.Google Scholar
  5. [5]
    D. Ball and S. Hoyt, The adaptive Time-Warp concurrency control algorithm,Proc. SCS Multiconf. on Distributed Simulation, Vol. 20 (1990) pp. 174–177.Google Scholar
  6. [6]
    S. Bellenot, Global virtual time algorithms,Proc. SCS Multiconf. on Distributed Simulation, Vol. 22 (1990) pp. 122–127.Google Scholar
  7. [7]
    S. Bellenot, State skipping performance with the Time Warp operating system, in:6th Workshop on Parallel and Distributed Simulation, Vol. 24, SCS Simulation Series (1992) pp. 53–64.Google Scholar
  8. [8]
    B. Berkman and R. Ayani, Parallel simulation of multistage interconnection networks on an SIMD computer, in:Advances in Parallel and Distributed Simulation, Vol. 23, SCS Simulation Series (1991) pp. 133–140.Google Scholar
  9. [9]
    J.V. Briner, Parallel mixed-level simulation of digital circuits using virtual time, Ph.D. Thesis, Duke University, Durham, NC (1990).Google Scholar
  10. [10]
    R.E. Bryant, Simulation of packet communication architecture computer systems, MIT-LCS-TR-188, Massachusetts Institute of Technology (1977).Google Scholar
  11. [11]
    C.A. Buzzell, M.J. Robb and R.M. Fujimoto, Modular VME rollback hardware for Time Warp,Proc. SCS Multiconf. on Distributed Simulation, Vol. 22 (1990) pp. 153–156.Google Scholar
  12. [12]
    W. Cai and S.J. Turner, An algorithm for distributed discrete-event simulation — the “carrier null message” approach,Proc. SCS Multiconf. on Distributed Simulation, Vol. 22 (1990) pp. 3–8.Google Scholar
  13. [13]
    K.M. Chandy, A survey of analytic models of rollback and recovery strategies, IEEE Comp. 8(1975)40–47.Google Scholar
  14. [14]
    K.M. Chandy and R. Sherman, The conditional event approach to distributed simulation,Proc. SCS Multiconf. on Distributed Simulation, Vol. 21 (1989) pp. 93–99.Google Scholar
  15. [15]
    K.M. Chandy and R. Sherman, Space, time, and simulation,Proc. SCS Multiconf. on Distributed Simulation, Vol. 21 (1989) pp. 53–57.Google Scholar
  16. [16]
    K.M. Chandy and J. Misra, Distributed simulation: A case study in design and verification of distributed programs, IEEE Trans. Software Eng. SE-5(1979)440–452.Google Scholar
  17. [17]
    A.I. Concepcion, A hierarchical computer architecture for distributed simulation, IEEE Trans. Comp. C-38(1989)311–319.Google Scholar
  18. [18]
    S.R. Das and R.M. Fujimoto, A performance study of the cancelback protocol for Time Warp, Technical Report GIT-CC-92/50, College of Computing, Georgia Institute of Technology, Atlanta, GA (1992).Google Scholar
  19. [19]
    M. Davoren, A structural mapping for parallel difital logic simulation, in:Proc. SCS Multiconf. on Distributed Simulation, Vol. 21, SCS Simulation Series (1989) pp. 179–182.Google Scholar
  20. [20]
    E. DeBenedictis, S. Ghosh and M.-L. Yu, A novel algorithm for discrete-event simulation, Computer 24(6)(1991)21–33.Google Scholar
  21. [21]
    P.M. Dickens, Performance analysis of parallel simulations, Ph.D. Thesis, University of Virginia (1992).Google Scholar
  22. [22]
    S. Eick, A. Greenberg, B. Lubachevsky and A. Weiss, Synchronous relaxation for parallel simulations with applications to circuit-switched networks, in:Advances in Parallel and Distributed Simulation, Vol. 23, SCS Simulation Series (1991) pp. 151–162.Google Scholar
  23. [23]
    R. Felderman and L. Kleinrock, An upper bound on the improvement of asynchronous versus synchronous distributed processing, in:Distributed Simulation, Vol. 22, SCS Simulation Series (1990) pp. 131–136.Google Scholar
  24. [24]
    R. Felderman and L. Kleinrock, Bounds and approximations for self-initiating distributed simulation without lookahead, ACM Trans. Modeling Comp. Simul. 1(1991).Google Scholar
  25. [25]
    R. Felderman and L. Kleinrock, Two processor Time Warp analysis: Some results on a unifying approach, in:Advances in Parallel and Distributed Simulation, Vol. 23, SCS Simulation Series (1991) pp. 3–10.Google Scholar
  26. [26]
    M.A. Franklin, D.F. Wann and K.F. Wong, Parallel machines and algorithms for discrete-event simulation,Proc. 1984 Int. Conf. on Parallel Processing (1984) pp. 449–458.Google Scholar
  27. [27]
    R.M. Fujimoto, Time Warp on a shared memory multiprocessor, Trans. Soc. Comp. Simul. 6(1989)211–239.Google Scholar
  28. [28]
    R.M. Fujimoto, The virtual time machine,Int. Symp. on Parallel Algorithms and Architectures (1989) pp. 199–208.Google Scholar
  29. [29]
    R.M. Fujimoto, Parallel discrete event simulation, Commun. ACM 33(1990)30–53.Google Scholar
  30. [30]
    R.M. Fujimoto, J. Tsai and G. Gopalakrishnan, Design and evaluation of the rollback chip: Special purpose hardware for Time Warp, IEEE Trans. Comp. C-41(1992)68–82.Google Scholar
  31. [31]
    A. Gafni, Rollback mechanisms for optimistic distributed simulation systems,Proc. SCS Multiconf. on Distributed Simulation, Vol. 19 (1988) pp. 61–67.Google Scholar
  32. [32]
    B. Gaujal, A. Greenberg and D. Nicol, A sweep algorithm for massively parallel simulation of circuit-switched networks, ICASE Technical Report ICASE-92-30 (1992), to appear in J. Parallel Distr. Comp.Google Scholar
  33. [33]
    E. Gelenbe, On the optimum checkpoint interval, J. ACM 26(1979)259–270.Google Scholar
  34. [34]
    P.I. Georgiadis, M.P. Papazoglou and D.G. Maritsas, Towards a parallel simula machine,Proc. 8th Annual Symp. on Computer Architecture, Vol. 9 (1982) pp. 263–278.Google Scholar
  35. [35]
    K. Ghosh and R.M. Fujimoto, Parallel discrete event simulation using space-time memory,Proc. 1991 Int. Conf. on Parallel Processing, Vol. 3 (1991) pp. 201–208.Google Scholar
  36. [36]
    D.W. Glazer, Load balancing parallel discrete-event simulation, Ph.D. Thesis, McGill University (1992).Google Scholar
  37. [37]
    A. Goldberg, Virtual time synchronization of replicated processes, in:6th Workshop on Parallel and Distributed Simulation, Vol. 24, SCS Simulation Series (1992) pp. 107–116.Google Scholar
  38. [38]
    G. Gopalakrishnan and R.M. Fujimoto, Design and verification of the rollback chip using HOP: A case study of formal methods applied to hardware design, Technical Report UUCS-91-015, Department of Computer Science, University of Utah (1991).Google Scholar
  39. [39]
    A.G. Greenberg, B.D. Lubachevsky and I. Mitrani, Algorithms for unboundedly parallel simulations, ACM Trans. Comp. Syst. 9(1991)201–221.Google Scholar
  40. [40]
    A. Gupta, I.F. Akyildiz and R.M. Fujimoto, Performance analysis of Time Warp with multiple homogeneous processors, IEEE Trans. Software Eng. SE-17(1991)1013–1027.Google Scholar
  41. [41]
    P. Heidelberger and D. Nicol, Conservative parallel simulation of continuous time Markov chains using uniformization, IBM Technical Report RC-16780, IBM Research Division (1991), to appear in IEEE Trans. Parallel Distr. Syst.Google Scholar
  42. [42]
    P. Heidelberger and H. Stone, Parallel trace-driven cache simulation by time partitioning, IBM Technical Report RC-15500, IBM Research Division (1990).Google Scholar
  43. [43]
    D.R. Jefferson, Virtual time, ACM Trans. Progr. Languages Syst. 7(1985)404–425.Google Scholar
  44. [44]
    D.R. Jefferson, Virtual time II: Storage management in distributed simulation,Proc. 9th Annual ACM Symp. on Principles of Distributed Computing (1990) pp. 75–89.Google Scholar
  45. [45]
    D.S. Johnson, C.R. Aragon, L.A. McGeoch and C. Shevon, Optimization by simulated annealing: An experimental evaluation: Part I, graph partitioning, Oper. Res. 37(1989)865–892.Google Scholar
  46. [46]
    S.A. Kravitz and B.D. Ackland, Static vs. dynamic partitioning of circuits for a MOS timing simulator on a message-based multiprocessor, in:Proc. SCS Multiconf. on Distributed Simulation, Vol. 19, SCS Simulation Series (1988) pp. 136–140.Google Scholar
  47. [47]
    D. Kumar and S. Harous, An approach towards distributed simulation of timed Petri nets, in:Proc. 1990 Winter Simulation Conf., New Orleans, LA (1990) pp. 428–435.Google Scholar
  48. [48]
    F. Lin and R.M. Keller, The gradient model oad balancing method, IEEE Trans. Software Eng. SE-11(1987)32–38.Google Scholar
  49. [49]
    Y.-B. Lin, Memory management algorithms for optimistic parallel simulation, in:6th Workshop on Parallel and Distributed Simulation, Vol. 24, SCS Simulation Series (1992).Google Scholar
  50. [50]
    Y.-B. Lin and E.D. Lazowska, Optimality considerations of “Time Warp” parallel simulation, in:Distributed Simulation, Vol. 22, SCS Simulation Series (1990) pp. 29–34.Google Scholar
  51. [51]
    Y.-B. Lin and E.D. Lazowska, Determining the global virtual time in a distributed simulation, Technical Report 90-01-02, Department of Computer Science, University of Washington, Seattle, WA (1989).Google Scholar
  52. [52]
    Y.-B. Lin and E.D. Lazowska, Reducing the state saving overhead for Time Warp parallel simulation, Technical Report 90-02-03, Department of Computer Science, University of Washington, Seattle, WA (1990).Google Scholar
  53. [53]
    Y.-B. Lin and E.D. Lazowska, A time-division algorithm for parallel simulation, ACM Trans. Mod. Comp. Simul. 1(1991)73–83.Google Scholar
  54. [54]
    Y.-B. Lin, E.D. Lazowska and J.-L. Baer, Conservative parallel simulation for systems with no lookahead prediction,Proc. SCS Multiconf. on Distributed Simulation, Vol. 22 (1990) pp. 144–149.Google Scholar
  55. [55]
    Y.-B. Lin and E.D. Lazowska, A study of Time Warp mechanisms, ACM Trans. Mod. Comp. Simul. 1(1991)51–72.Google Scholar
  56. [56]
    Y.-B. Lin and E.D. Lazowska, A time-division algorithm for parallel simulation, ACM Trans. Mod. Comp. Simul. 1(1991)73–83.Google Scholar
  57. [57]
    R. Lipton and D. Mizell, Time Warp vs. Chandy-Misra: A worst-case comparison, in:Distributed Simulation, Vol. 22, SCS Simulation Series (1990) pp. 137–143.Google Scholar
  58. [58]
    B. Lubachevsky, A. Weiss and A. Shwartz, An analysis of rollback-based simulation, ACM Trans. Mod. Comp. Simul. 1(1991)154–192.Google Scholar
  59. [59]
    B.D. Lubachevsky, Scalability of the bounded lag distributed discrete event simulation,Proc. SCS Multiconf. on Distributed Simulation, Vol. 21 (1989) pp. 100–107.Google Scholar
  60. [60]
    B.D. Lubachevsky, A. Shwartz and A. Weiss, Rollback sometimes works ... if filtered,1989 Winter Simulation Conf. Proc. (1989) pp. 630–639.Google Scholar
  61. [61]
    V. Madissetti, D. Hardaker and R. Fujimoto, The mimdix operating system for parallel simulation, in:6th Workshop on Parallel and Distributed Simulation, Vol. 24, SCS Simulation Series (1992) pp. 65–74.Google Scholar
  62. [62]
    V. Madisetti, J. Walrand and D. Messerschmitt, Wolf: A rollback algorithm for optimistic distributed simulation systems,1988 Winter Simulation Conf. Proc. (1988) pp. 296–305.Google Scholar
  63. [63]
    J. Briner, Jr., Fast parallel simulation of digital systems, in:Advances in Parallel and Distributed Simulation, Vol. 23, SCS Simulation Series (1991) pp. 71–77.Google Scholar
  64. [64]
    P. Reynolds, Jr., An efficient framework for parallel simulations, in:Advances in Parallel and Distributed Simulation, Vol. 23, SCS Simulation Series (1991) pp. 167–174.Google Scholar
  65. [65]
    J. Misra, Distributed discrete-event simulation, ACM Comp. Surveys 18(1986)39–65.Google Scholar
  66. [66]
    B. Nandy and W. Loucks, An algorithm for partitioning and mapping conservative parallel simulation onto multicomputers, in:6th Workshop on Parallel and Distributed Simulation, Vol. 24, SCS Simulation Series (1992) pp. 139–146.Google Scholar
  67. [67]
    L.M. Ni, C.W. Zu and T.B. Gendreau, A distributed drafting algorithm for load balancing, IEEE Trans. Software Eng. SE-9(1985).Google Scholar
  68. [68]
    D. Nicol, Optimistic barrier synchronization, ICASE Technical Report 91–34 (1992).Google Scholar
  69. [69]
    D. Nicol, A. Greenberg, B. Lubachevsky and S. Roy, Massively parallel algorithms for trace-driven cache simulation, in:6th Workshop on Parallel and Distributed Simulation, Vol. 24, SCS Simulation Series (1992) pp. 3–11.Google Scholar
  70. [70]
    D. Nicol and P. Heidelberger, Optimistic parallel simulation of continuous time Markov chains using uniformization, J. Parallel Distr. Comp. 18(1993)395–410.Google Scholar
  71. [71]
    D. Nicol and P. Heidelberger, Parallel simulation of Markovian queueing networks using adaptive uniformization, in:Proc. 1993 SIGMETRICS Conf, Santa Clara, CA (1993) pp. 135–145.Google Scholar
  72. [72]
    D. Nicol and S. Roy, Parallel simulation of timed Petri nets, in:Proc. 1991 Winter Simulation Conf., Phoenix, AZ (1991) pp. 574–583.Google Scholar
  73. [73]
    D.M. Nicol, Performance bounds on parallel self-initiating discrete-event simulations, ACM Trans. Mod. Comp. Simul. 1(1991)24–50.Google Scholar
  74. [74]
    D.M. Nicol, The automated partitioning of simulations for parallel execution, Ph.D. Thesis, University of Virginia (1985).Google Scholar
  75. [75]
    D.M. Nicol, The cost of conservative synchronization in parallel discrete-event simulations, J. ACM 40(1993)304–333.Google Scholar
  76. [76]
    D.M. Nicol and P.F. Reynolds, Jr., A statistical approach to dynamic partitioning, in:Distributed Simulation 85, Vol. 15, SCS Simulation Series (1985) pp. 53–56.Google Scholar
  77. [77]
    D.M. Nicol and P.F. Reynolds, Jr., Optimal dynamic remapping of data parallel computations, IEEE Trans. Comp. C-39(1990)206–219.Google Scholar
  78. [78]
    C. Pancerella, Improving the efficiency of a framework for parallel simulations, in:6th Workshop on Parallel and Distributed Simulation, Vol. 24, SCS Simulation Series (1992) pp. 22–32.Google Scholar
  79. [79]
    B. Preiss, W. Loucks, I. MacIntyre and J. Field, Null message cancellation in conservative distributed simulation, in:Advances in Parallel and Distributed Simulation, Vol. 23, SCS Simulation Series (1991) pp. 33–38.Google Scholar
  80. [80]
    B. Preiss, I. MacIntyre and W. Loucks, On the trade-off between time and space in optimistic parallel discrete-event simulation, in:6th Workshop on Parallel and Distributed Simulation, Vol. 24, SCS Simulation Series (1992) pp. 33–42.Google Scholar
  81. [81]
    B.R. Preiss, The Yaddes distributed discrete event simulation specification language and execution environments,Proc. SCS Multiconf. on Distributed Simulation, Vol. 21 (1989) pp. 139–144.Google Scholar
  82. [82]
    P.L. Reiher and D. Jefferson, Dynamic load management in the Time Warp Operating System, Trans. Soc. Comp. Simul. 7(1990)91–120.Google Scholar
  83. [83]
    H.R. Ross,Stochastic Processes (Wiley, New York, 1983).Google Scholar
  84. [84]
    B. Samadi, Distributed simulation, algorithms and performance analysis, Ph.D. Thesis, University of California, Los Angeles (1985).Google Scholar
  85. [85]
    L.M. Sokol, D.P. Briscoe and A.P. Wieland, MTW: a strategy for scheduling discrete simulation events for concurrent execution,Proc. SCS Multiconf. on Distributed Simulation, Vol. 19 (1988) pp. 34–42.Google Scholar
  86. [86]
    L. Soule and A. Gupta, An evaluation of the Chandy-Misra-Byrant algorithm for digital logic simulation, ACM Trans. Mod. Comp. Simul. 1 (1991).Google Scholar
  87. [87]
    J. Steinman, Speedes: synchronous parallel environment for emulation and discrete event simulation, in:Advances in Parallel and Distributed Simulation, Vol. 23, SCS Simulation Series (1991) pp. 95–103.Google Scholar
  88. [88]
    W.K. Su and C.L. Seitz, Variants of the Chandy-Misra-Bryant distributed discrete-event simulation algorithm,Proc. SCS Multiconf. on Distributed Simulation, Vol. 21 (1989) pp. 38–43.Google Scholar
  89. [89]
    G. Thomas and J. Zahorjan, Parallel simulation of performance Petra nets: Extending the domain of parallel simulation, in:Proc. 1991 Winter Simulation Conf., Phoenix, AZ (1991) pp. 564–573.Google Scholar
  90. [90]
    S. Turner and M. Xu, Performance evaluation of the bounded Time Warp algorithm, in:6th Workshop on Parallel and Distributed Simulation, Vol. 24, SCS Simulation Series (1992) pp. 117–128.Google Scholar

Copyright information

© J.C. Baltzer AG, Science Publishers 1994

Authors and Affiliations

  • David Nicol
    • 1
  • Richard Fujimoto
    • 2
  1. 1.Department of Computer ScienceCollege of William and MaryWilliamsburgUSA
  2. 2.College of ComputingGeorgia Institute of TechnologyAtlantaUSA

Personalised recommendations