IBM AWM Course Content :

1. Introduction to Application Workload Modeler (AWM)

• Motivation
• Addressed Types of Systems
• Contribution of the Presented Work

2. Motivation

• Open Problems in Performance Evaluation.
• Demands on a workload modeling framework and methodology for Parallel Processing Systems.

3. Addressed Types of Systems

• User System Interaction
• System under Test and Component under Study
• Parallel architectures.
• Parallel Programs

4. Contribution of the Presented Work

• Summary of the Framework and the Methodology.
• Outline of the Thesis

5. Workload Modeling –Framework and Methodology

• Definition of Workload
• Workload Model Construction in a Performance Evaluation Cycle
• Constructing a Non- Executable Test Workload Model

6. Definition of Workload

• Workload Components
• Specifying the Test Workload
• Parallel Workload Characteristics

7. Workload Model Construction in a Performance Evaluation Cycle

• Influence of the Objective of the Evaluation Study
• Influence of the Evaluation Techniques

8. Constructing a Non- Executable Test Workload Model

• Model Parameterization
• Validation Criteria for Workload Models

9. Survey and Comparison of Parallel Workload Models

• Parameter Based Models
• Profiles, Signatures and Shapes
• Ratio’s for Varying System Size
• Graph Models

10. Parameter Based Models

11. Profiles, Signatures and Shapes

12. Ratio’s for Varying System Size

13. Graph Models

• Undirected Graph Models
• Directed Graph Models

14. Scalability Analysis

• Scalability Concepts
• Visual Approach for Scalability Analysis
• Examples

15. Scalability Concepts

• Problem Definitions
• Scalability Index
• Characterizing the Architecture Size A
• Characterizing the Amount of Work W
• Sources of Overhead
• Evaluation Techniques for Scalability Analysis

16. Visual Approach for Scalability Analysis

• Performance Characteristics
• Model Parameters and Definitions
• Comparison to other Approaches

17. Examples

• An Embarrassingly Parallel Problem
• Householder Reduction
• Finite Element Method

18. Summary and Future Work

• Abbreviations and Symbols
• Definitions and Derivation

19. Abbreviations and Symbols

20. Definitions and Derivation

• Petri Nets
• Series Parallel Graphs
• Stochastic Process
• Queuing Networks
• Complexity Notation
• Bounds on and Relations between Ratios’s
• Bounds