Visualizing structural properties of irregular parallel computations

Authors:
Wolfgang Blochinger;Michael Kaufmann;Martin Siebenhaller
Affiliations:
University of Tübingen, Germany;University of Tübingen, Germany;University of Tübingen, Germany
Venue:
SoftVis '05 Proceedings of the 2005 ACM symposium on Software visualization
Year:
2005

Citing 13
Cited 0

Space-efficient scheduling of multithreaded computations

STOC '93 Proceedings of the twenty-fifth annual ACM symposium on Theory of computing
UML distilled (2nd ed.): a brief guide to the standard object modeling language

UML distilled (2nd ed.): a brief guide to the standard object modeling language
An object-oriented platform for distributed high-performance symbolic computation

Mathematics and Computers in Simulation - Special issue on high performance symbolic computing
Layered drawings of digraphs

Drawing graphs
Creating an Accurate Portrayal of Concurrent Executions

IEEE Concurrency
The Paradyn Parallel Performance Measurement Tool

Computer
Visualizing the Performance of Parallel Programs

IEEE Software
Upper Bounds on the Number of Hidden Nodes in Sugiyama's Algorithm

GD '96 Proceedings of the Symposium on Graph Drawing
Fast and Simple Horizontal Coordinate Assignment

GD '01 Revised Papers from the 9th International Symposium on Graph Drawing
The Java Developer's Guide to Eclipse

The Java Developer's Guide to Eclipse
Parallel propositional satisfiability checking with distributed dynamic learning

Parallel Computing
Cilk: efficient multithreaded computing

Cilk: efficient multithreaded computing
An efficient implementation of sugiyama's algorithm for layered graph drawing

GD'04 Proceedings of the 12th international conference on Graph Drawing

Quantified Score

Hi-index	0.00

Visualization

Abstract

An important task in parallel programming is the appropriate distribution of work on the processors. This distribution is usually dynamically changing and hard to predict, further it is very sensitive to the change of parameters. Even with advanced analysis tools this problem is hard to be solved. We propose to visualize the program structure as it changes over the execution time. We therefore present a new automatic layout algorithm based on Sugiyama's framework, which enables the user to detect structural patterns which might be fatal for the performance of the program - patterns which might be impossible to detect in a more analytical way. Furthermore it assists the user to find appropriate timing parameters for load balancing. We integrate our visualization into an integrated development environment that supports the implementation, execution, and analysis of parallel programs.