PERFORMANCE FACTORS IN PARALLEL DISCRETE EVENT SIMULATION

Performance is a key issue in parallel simulation. This paper proposes a standard approach for detailed performance analysis of parallel discrete event simulation. We developed it for our conservative simulation algorithm, but it can serve for parallelisation in general as well. The parallel overhead terms are introduced and model-dependency of the parallel speedup is expressed by the performance factors. Experimental results show that this analysis can serve for speedup bottleneck detection. We propose this approach as a workplan for in-depth investigation of the performance (performance prediction, scalability analysis and algorithm comparison).

[1]  Wentong Cai,et al.  Performance prediction tools for parallel discrete-event simulation , 1999, Proceedings Thirteenth Workshop on Parallel and Distributed Simulation. PADS 99. (Cat. No.PR00155).

[2]  Randal E. Bryant,et al.  SIMULATION OF PACKET COMMUNICATION ARCHITECTURE COMPUTER SYSTEMS , 1977 .

[3]  K. Mani Chandy,et al.  Distributed Simulation: A Case Study in Design and Verification of Distributed Programs , 1979, IEEE Transactions on Software Engineering.

[4]  Roger D. Chamberlain,et al.  Exploiting lookahead in synchronous parallel simulation , 1993, WSC '93.

[5]  Alois Ferscha,et al.  A testbed for parallel simulation performance prediction , 1996, Winter Simulation Conference.

[6]  George Karypis,et al.  Introduction to Parallel Computing , 1994 .

[7]  David M. Nicol,et al.  Performance prediction of a parallel simulator , 1999, Proceedings Thirteenth Workshop on Parallel and Distributed Simulation. PADS 99. (Cat. No.PR00155).

[8]  Richard M. Fujimoto,et al.  Feature Article - Parallel Discrete Event Simulation: Will the Field Survive? , 1993, INFORMS J. Comput..

[9]  D. Nicol.  Scalability, locality, partitioning and synchronization in PDES , 1998, Proceedings. Twelfth Workshop on Parallel and Distributed Simulation PADS '98 (Cat. No.98TB100233).

[10]  Satish K. Tripathi,et al.  Parallel and distributed simulation of discrete event systems , 1994 .

[11]  R. Bagrodia,et al.  A Performance Evaluation Methodology for Parallel Simulation Protocols , 1996, Proceedings of Symposium on Parallel and Distributed Tools.

[12]  Eunmi Choi,et al.  An important factor for optimistic protocol on distributed systems: granularity , 1995, Winter Simulation Conference Proceedings, 1995..

[13]  R. M. Fujimoto,et al.  Parallel discrete event simulation , 1989, WSC '89.