并行程序设计(英文版)
分类: 图书,计算机/网络,计算机理论,
作者: 福斯特编著
出 版 社: 人民邮电出版社
出版时间: 2002-9-1字数: 447千版次: 第1版第1次页数: 381印刷时间: 2002/09/01开本:印次:纸张: 胶版纸I S B N : 9787115103475包装: 平装编辑推荐
内容简介
本书讲解如何在并行和分布式操作系统下设计高效率、低开销的程序,内容既包括并行程序设计的技术也包括实现程序的工具。
全书分为三个部分13章。第一部分是前4章,介绍并行计算和程序设计的概念。第二部分介绍了并行程序设计的语言和函数库,包括C++、Fortran M、HPF和MPI等进行编程工具。第三部分给出了并行程序的几类算法和常用的资源。
本书适合作为高等学院校计算机专业并行程序设计课程的教材,也适合具有相应水平的读者自学。
作者简介
目录
PARTⅠCONCEPTS1
1Parallel Computers and Computation3
1.1Parallelism and Computing3
1.2A Parallel Machine Model9
1.3A Parallel Programming Model12
1.4Parallel Algorithm Examples18
1.5Summary23
Exercises23
Chapter Notes24
2Designing Parallel Algorithms27
2.1Methodical Design27
2.2Partitioning28
2.3Communication32
2.4Agglomeration42
2.5Mapping50
2.6Case Study: Atmosphere Model57
2.7Case Study: Floorplan Optimization64
2.8Case Study: Computational Chemistry71
2.9Summary77
Exercises78
Chapter Notes80
3A Quantitative Basis for Design83
3.1Defining Performance83
3.2Approaches to Performance Modeling84
3.3Developing Models88
3.4Scalability Analysis95
3.5Experimental Studies100
3.6Evaluating Implementations106
3.7A Refined Communication Cost Model111
3.8Input/Output121
3.9Case Study: Shortest-Path Algorithms123
3.10Summary130
Exercises130
Chapter Notes133
4Putting Components Together137
4.1Modular Design Review137
4.2Modularity and Parallel Computing139
4.3Performance Analysis146
4.4Case Study: Convolution147
4.5Case Study: Tuple Space152
4.6Case Study: Matrix Multiplication155
4.7Summary159
Exercises161
Chapter Notes162
PARTⅡTOOLS165
5Compositional C++167
5.1C++ Review167
5.2CC++ Introduction171
5.3Concurrency172
5.4Locality174
5.5Communication176
5.6Asynchronous Communication186
5.7Determinism187
5.8Mapping187
5.9Modularity193
5.10Performance Issues196
5.11Case Study: Channel Library200
5.12Case Study: Fock Matrix Construction201
5.13Summary202
Exercises203
Chapter Notes204
6Fortran M207
6.1FM Introduction207
6.2Concurrency210
6.3Communication212
6.4Unstructured Communication216
6.5Asynchronous Communication222
6.6Determinism225
6.7Argument Passing226
6.8Mapping228
6.9Modularity232
6.10Performance Issues235
6.11Case Study: Fock Matrix Construction238
6.12Summary238
Exercises239
Chapter Notes242
7High Performance Fortran243
7.1Data Parallelism243
7.2Fortran 90246
7.3Data Distribution250
7.4Concurrency256
7.5Dummy Arguments and Modularity259
7.6Other HPF Features261
7.7Performance Issues264
7.8Case Study: Gaussian Elimination268
7.9Summary271
Exercises272
Chapter Notes273
8Message Passing Interface275
8.1The MPI Programming Model275
8.2MPI Basics276
8.3Global Operations283
8.4Asynchronous Communication290
8.5Modularity293
8.6Other MPI Features300
8.7Performance Issues304
8.8Case Study: Earth System Model305
8.9Summary305
Exercises310
Chapter Notes311
9Performance Tools313
9.1Performance Analysis313
9.2Data Collection315
9.3Data Transformation and Visualization318
9.4Tools320
9.5Summary324
Exercises325
Chapter Notes325
PART ⅢRESOURCES327
10Random Numbers329
10.1Sequential Random Numbers329
10.2Parallel Random Numbers330
10.3Distributed Random Generators331
10.4Summary334
Exercises334
Chapter Notes335
11Hypercube Algorithms337
11.1The Hypercube Template337
11.2Vector Reduction338
11.3Matrix Transposition340
11.4Mergesort341
11.5Summary346
Exercises346
Chapter Notes347
12Further Reading349
13About the Online Version351
13.1Obtaining a Web Browser352
13.2Web Tours352
References353
Index369
媒体评论