CS/CSE 6236 and CS 4230

Parallel and Distributed Simulation

Fall Semester, 2007

 

IMPORTANT: Current course information is being maintained at T-Square. Please refer to that web site for information concerning this course.

 

https://t-square.gatech.edu

 

 

 

Course Syllabus and Schedule

 

 

Week

Date

Tuesday

Thursday

Reading Assignment

1

8/21 - 8/23

Parallel/Distributed Simulation

Discrete Event Simulation

Chapters 1 and 2

2

8/28 - 8/30

Time Warp - Local Control

Time Warp - Global Control

4.1-4.4, 5.4

3

9/4 - 9/6

Real-Time Distributed Simulation

TBD

 

4

9/11 - 9/13

GVT Algorithms (cont.)

Reverse execution

[Reverse]

5

9/18 - 9/20

Ad Hoc Distributed Simulation

Simulation Cloning

[Ad Hoc][Cloning]

6

9/25 - 9/27

Time Warp state saving

Time Warp Memory Management

5.1-5.3

7

10/2 - 10/4

Aurora

Exam I

[Aurora]

8

10/9 - 10/11

no class - Fall break

Conservative execution: Null messages

3.1-3.2

9

10/16 - 10/18

Lookahead

Deadlock Detection & Recovery

3.3-3.4

10

10/23 - 10/25

Synchronous algorithms

Synch. Algorithms (cont.)

3.5-3.12

11

10/30 - 11/1

High Level Architecture

HLA Time management

7.6 [HLA-TM]

12

11/6 - 11/8

HLA Lookahead

Exam II

 

13

11/13 - 11/15

Time Parallel Algorithms

Time Parallel Algorithms (cont.);

Chapter 6

14

11/20 - 11/22

Distributed Virtual Environments

Thanksgiving Holiday

7.1-7.4

15

11/27 - 11/29

Data distribution

Data distribution (cont.)

8.5-8.6

16

12/4 - 12/6

Dead Reckonning

Clock Synchronization

7.5, 9.3

 

12/12/07

Final Exam: 11:30 AM - 2:20 PM

 

 

 

 

 

 

 

 

Drop Date: 10/12/07  

Instructor

Richard Fujimoto, fujimoto (at) cc (dot) gatech (dot) edu  404/894-5615

Office hours: TuTh 10:00-11:30 AM or by appointment (1320A KACB)

Teaching Assistant

TBD

Office hours: TBD

Class Time and Location

TuTh 1:35-2:55 PM, 2247 Klaus Advanced Computing Building

Textbook

R. Fujimoto, Parallel and Distributed Simulation Systems. John Wiley & Sons, 2000 (required).

Additional Readings:

[Ad Hoc] R. M. Fujimoto, M. Hunter, J. Sirichoke, M. Palekar, H.-K. Kim, W. Suh, “Ad Hoc Distributed Simulations,” Principles of Advanced and Distributed Simulation, June 2007 [pdf].

[Aurora] A. Park, R. M. Fujimoto, “Aurora: An Approach to High Throughput Parallel Simulation,” Principles of Advanced and Distributed Simulation, May 2006 [pdf].

[Cloning] M. Hybinette and R. M. Fujimoto “Cloning Parallel Simulations,” ACM Transactions on Modeling and Computer Simulation, Vol. 11, No. 3, pp. 378-407, October 2001 [pdf].

[HLA-TM] R. M. Fujimoto, “Time Management in the High Level Architecture,” Simulation, Vol. 71, No. 6, pp. 388-400, December 1998 [pdf]

[Reverse] C. D. Carothers, K. Perumalla, R. M. Fujimoto, "Efficient Optimistic Parallel Simulation Using Reverse Computation, ACM Transactions on Modeling and Computer Simulation, Vol. 9, No. 3, pp. 224-253, October 1999 [pdf].

Prerequisites

A general background in operating systems (processes, concurrency issues, semaphores, etc.) is assumed.  An undergraduate course in operating systems should suffice.  Programming assignments and projects will require software development in C and/or C++.  No prior knowledge of simulation or parallel or distributed computing is assumed. 

Course Requirements

Required work includes (1) homework assignments (2) a course project, (3) two in-class exams and a final examination. Each of the exams will count for 20% of the final grade, the project will count for 30%, and the homework count for the remaining 10%.  A list of possible projects will be distributed, or you are free to come up with your own (which must be approved in advance).

Computer accounts will be needed to complete the homework assignments and project. You should have access to the CoC Cluster and various desktop machines.

Course Content

This course is concerned with the realization of simulations on parallel or distributed computing systems for analytic or virtual environment applications.  The emphasis will be on discrete event simulations. After some introductory materials, the course is roughly divided into two parts.  The first is concerned with parallel and distributed execution of analytic simulations, e.g., for system evaluation.  The second is concerned with the execution of simulations on distributed computers to create virtual environments, e.g., for training or gaming applications.

Contact Information:

fujimoto (at) cc (dot) gatech (dot) edu
College of Computing
Georgia Institute of Technology
Atlanta, GA 30332-0280