MetaCartSign in to MyCiteSeer

Include Citations | Advanced Search | Help

Include Citations | Advanced Search | Help

  Parallel VHDL Simulation Engine

Download:
pdf
by Moon Jung Chung, Jinsheng Xu, Hee Chul Kim
http://www.cse.msu.edu/~xujinshe/pvhdl.pdf
Add To MetaCart

Abstract:

In this paper, we report a parallel simulation engine and its benchmark results on CEN3-Project "Parallel VHDL Performance Simulation". The goal of the project is to develop a parallel event driven simulator that will be able to simulate a model with up-to 100,000 concurrently interacting objects. Our target is to achieve a speed-up of up-to hundred times when implemented on a massively parallel system. We have developed a simulation framework that can be used to simulate not only VHDL models but also models in other application domains such as networks, force modeling, and traffic simulation. The simulation framework consists of the cockpit, the object modeling scheme, and the front end user interface. The cockpit, frequently called a simulation engine, is the main program module of the parallel simulator. It orchestrates modules, message handling, and event queue based on the simulation protocol it uses. The object modeling-scheme is developed using C++. Our modeling-scheme enables the modelers to develop simulation models independent from the simulation protocols. The end user builds the model of the system to be simulated by instantiating the components of the system, and providing their interconnection, and specifying the employed simulation protocols. We have benchmarked the performance of our simulator with various test suites on SP2 and Origin2000. The test suites have models consisting of more than 100,000 objects as their components. We have analyzed the communication patterns of parallel simulation and developed a mathematical model to predict the performance of parallel simulation in terms of the problem size, the activity rate, roll back rate, and the number of processors. We have observed that the performance of parallel simulation depends on computational granularity of the system to be simulated and the event scheduling. 1.

Citations

694 Virtual time – Jefferson - 1985
81 BP: Object-oriented simulation with hierarchical, modular models – Zeigler
41 The Yaddes distributed discrete event simulation specification language and execution environments – Preiss - 1989
21 Moose: An objectoriented multimodeling and simulation application framework, Simulation – Cuburt, Fishwick - 1997
16 Distributed discrete-event simulation. Computing Surveys – Misra - 1986
11 Scalable Parallel And Distributed Military Simulations Using the SPEEDES Framework – Steinman - 1995
8 Comparing Synchronization Protocols for Parallel Logic-Level Simulation – Lin, Lazowska, et al. - 1990
7 Actor based parallel VHDL simulation using Time Warp – Krishnaswamy, Banerjee - 1996
7 Dynamic model abstraction – Lee, Fishwick - 1996
5 A visual object-oriented multimodeling design approach for physical modeling – Fishwick - 1997
4 an environment for parallel discrete-event simulation – Sima - 1992
2 Efficient Parallel Logic Simulation Techniques for the Connection Machine – Chung, Chung - 1990
1 Designing Efficient Simulations Using Maisie, UCLA – Bagrodia - 1997
1 A Comparative Analysis of Various Time – Radhakrishnan, McBrayer, et al.
1 Chung and Jiashen Zhou, "Version Control and Configuration Management in Simulation – Jung - 1998