Interactive debugging and performance analysis of massively parallel applications

Wismüller, Roland; Oberhuber, Michael; Krammer, J.; Hansen, Olav R.

doi:10.1016/0167-8191(95)00070-4

Search citation statements

Order By: Relevance

Paper Sections

Select...

Introduction2

Citation Types

Supporting

Mentioning

Contrasting

Year Published

2000

2003

Publication Types

Select...

Book5

Article1

Relationship

Self Cite0

Independent6

Authors

Journals

Cited by 20 publications

(2 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Traditional debuggers are not particularly helpful at finding errors introduced during the porting process even when they have been designed with scientific or distributed computing in mind [6,8,15,16,19,25]. This is because they generally require the programmer to have a good understanding of the way the program works and have a mental model of the contents of the various data structures during execution.…”

Section: Introductionmentioning

confidence: 99%

Debugging scientific applications in the .NET Framework

Abramson

Watson

2003

Future Generation Computer Systems

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 99%

Debugging scientific applications in the .NET Framework

Abramson

Watson

2003

Future Generation Computer Systems

View full text Add to dashboard Cite

“…This seems almost archaic compared to the functionality provided by sophisticated debugging tools, such as DDBG [4], P2D2 [7], DETOP [13], and Totalview [5].…”

Section: Introductionmentioning

confidence: 99%

Parallel Program Debugging with MAD — A Practical Approach

Kranzlmüller¹,

Rimnac

2003

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Debugging parallel programs can be very time-consuming and tedious, because the multiplicity of communicating processes increases the complexity of a program and the probability of incorrect behavior. Solutions are provided by debugging tools, which try to offer meaningful ways to investigate errors and their original causes. The MAD environment is a debugging toolset, which focuses on parallel and distributed programs. This paper discusses the application of MAD to real-world programs with practicability and usability as the main goals. The observations are based on a series of debugging sessions conducted for a specific application of theoretical physics, with the opinions of the tool developer on one side and the view of the application developer on the other side.

show abstract

Performance Prediction of an NAS Benchmark Program with ChronosMix Environment

Bourgeois

Spies

2000

Euro-Par 2000 Parallel Processing

View full text Add to dashboard Cite

Interactive debugging and performance analysis of massively parallel applications

Cited by 20 publications

References 6 publications

Debugging scientific applications in the .NET Framework

Debugging scientific applications in the .NET Framework

Parallel Program Debugging with MAD — A Practical Approach

Performance Prediction of an NAS Benchmark Program with ChronosMix Environment

Contact Info

Product

Resources

About