High Performance Heterogeneous Computing

by Jack Dongarra; Alexey L. Lastovetsky (Univ. College, Dublin, Ireland )

• ISBN13: 9780470508190
• ISBN10: 0470508191
• Format: eBook
• Copyright: 2009-08-01
• Publisher: Wiley-Interscience

Summary

An analytical overview of the state of the art, open problems, and future trends in heterogeneous parallel and distributed computingThis book provides an overview of the ongoing academic research, development, and uses of heterogeneous parallel and distributed computing in the context of scientific computing. Presenting the state of the art in this challenging and rapidly evolving area, the book is organized in five distinct parts:Heterogeneous Platforms: Taxonomy, Typical Uses, and Programming IssuesPerformance Models of Heterogeneous Platforms and Design of?Heterogeneous AlgorithmsPerformance: Implementation and SoftwareApplicationsFuture TrendsHigh Performance Heterogeneous Computing is a valuable?reference for researchers and practitioners in the area of high performance heterogeneous computing. It also serves as an excellent supplemental text for graduate and postgraduate courses in related areas.

Table of Contents

Preface

Acknowledgments

Heterogeneous Platforms: Taxonomy, Typical Uses, And Programming Issues

Heterogeneous Platforms and Their Uses

Taxonomy of Heterogeneous Platforms

Vendor-Designed Heterogeneous Systems

Heterogeneous Clusters

Local Network of Computers (LNC)

Global Network of Computers (GNC)

Grid-Based Systems

Other Heterogeneous Platforms

Typical Uses of Heterogeneous Platforms

Traditional Use

Parallel Computing

Distributed Computing

Programming Issues

Performance

Fault Tolerance

Arithmetic Heterogeneity

Performance Models Of Heterogeneous Platforms And Design Of Heterogeneous Algorithms

Distribution of Computations with Constant Performance Models of Heterogeneous Processors

Simplest Constant Performance Model of Heterogeneous Processors and Optimal Distribution of Independent Units of Computation with This Model

Data Distribution Problems with Constant Performance Models of Heterogeneous Processors

Partitioning Well-Ordered Sets with Constant Performance Models of Heterogeneous Processors

Partitioning Matrices with Constant Performance Models of Heterogeneous Processors

Distribution of Computations with Nonconstant Performance Models of Heterogeneous Processors

Functional Performance Model of Heterogeneous Processors

Data Partitioning with the Functional Performance Model of Heterogeneous Processors

Other Nonconstant Performance Models of Heterogeneous Processors

Stepwise Functional Model

Functional Model with Limits on Task Size

Band Model

Communication Performance Models for High-Performance Heterogeneous Platforms

Modeling the Communication Performance for Scientific Computing: The Scope of Interest

Communication Models for Parallel Computing on Heterogeneous Clusters

Communication Performance Models for Local and Global Networks of Computers

Performance Analysis of Heterogeneous Algorithms

Efficiency Analysis of Heterogeneous Algorithms

Scalability Analysis of Heterogeneous Algorithms

Performance: Implementation And Software

Implementation Issues

Portable Implementation of Heterogeneous Algorithms and Self-Adaptable Applications

Performance Models of Heterogeneous Platforms: Estimation of Parameters

Estimation of Constant Performance Models of Heterogeneous Processors

Estimation of Functional and Band Performance Models of Heterogeneous Processors

Benchmarking of Communication Operations

Performance Models of Heterogeneous Algorithms and Their Use in Applications and Programming Systems

Implementation of Homogeneous Algorithms for Heterogeneous Platforms

Programming Systems for High-Performance Heterogeneous Computing

Parallel Programming Systems for Heterogeneous Platforms

Traditional Parallel Programming Systems

Message-Passing Programming Systems

Linda

HPF

Heterogeneous Parallel Programming Systems

Distributed Programming Systems

NetSolve

Nimrod

Java

GridRPC

Applications

Numerical Linear Algebra Software for Heterogeneous Clusters

HeteroPBLAS: Introduction and User Interface

HeteroPBLAS: Software Design

Experiments with HeteroPBLAS

Parallel Processing of Remotely Sensed Hyperspectral Images on Heterogeneous Clusters

Hyperspectral Imaging: Introduction and Parallel Techniques

A Parallel Algorithm for Analysis of Hyperspectral Images and Its Implementation for Heterogeneous Clusters

Experiments with the Heterogeneous Hyperspectral Imaging Application

Conclusion

Simulation of the Evolution of Clusters of Galaxies on Heterogeneous Computational Grids

Hydropad: A Simulator of Galaxies' Evolution

Enabling Hydropad for Grid Computing

GridRPC Implementation of the Hydropad

Experiments with the GridSolve-Enabled Hydropad

SmartGridSolve and Hydropad

SmartGridSolve Implementation of the Hydropad

Experiments with the SmartGridSolve-Enabled Hydropad

Acknowledgment

Future Trends

Future Trends in Computing

Introduction

Computational Resources

Complex and Heterogeneous Parallel Systems

Intel-ization of the Processor Landscape

New Architectures on the Horizon

Applications

Software

Some Important Concepts for the Future

Heterogeneous Hardware Environments

Software Architecture

Open Source

New Applications

Verification and Validation

Data

2009 and Beyond

References

Appendices

Appendix to Chapter 3

Proof of Proposition 3.1

Proof of Proposition 3.5

Appendix to Chapter 4

Proof of Proposition 4.1

Proof of Proposition 4.2

Proof of Proposition 4.3

Functional Optimization Problem with Optimal Solution, Locally Nonoptimal

Index

Table of Contents provided by Publisher. All Rights Reserved.