Muutke küpsiste eelistusi

Parallel Science and Engineering Applications: The Charmplusplus Approach [Kõva köide]

Edited by , Edited by
  • Formaat: Hardback, 314 pages, kõrgus x laius: 234x156 mm, kaal: 589 g
  • Ilmumisaeg: 28-Oct-2013
  • Kirjastus: CRC Press Inc
  • ISBN-10: 1466504129
  • ISBN-13: 9781466504127
Teised raamatud teemal:
Parallel Science and Engineering Applications: The Charmplusplus ApproachSuurem pilt
  • Formaat: Hardback, 314 pages, kõrgus x laius: 234x156 mm, kaal: 589 g
  • Ilmumisaeg: 28-Oct-2013
  • Kirjastus: CRC Press Inc
  • ISBN-10: 1466504129
  • ISBN-13: 9781466504127
Teised raamatud teemal:
Püsilink: https://www.kriso.ee/db/9781466504127.html
Märksõnad:
Developed in the context of science and engineering applications, with each abstraction motivated by and further honed by specific application needs, Charm++ is a production-quality system that runs on almost all parallel computers available. Parallel Science and Engineering Applications: The Charm++ Approach surveys a diverse and scalable collection of science and engineering applications, most of which are used regularly on supercomputers by scientists to further their research.





After a brief introduction to Charm++, the book presents several parallel CSE codes written in the Charm++ model, along with their underlying scientific and numerical formulations, explaining their parallelization strategies and parallel performance. These chapters demonstrate the versatility of Charm++ and its utility for a wide variety of applications, including molecular dynamics, cosmology, quantum chemistry, fracture simulations, agent-based simulations, and weather modeling.





The book is intended for a wide audience of people in academia and industry associated with the field of high performance computing. Application developers and users will find this book interesting as an introduction to Charm++ and to developing parallel applications in an asynchronous message-driven model. It will also be a useful reference for undergraduate and graduate courses in computer science and other engineering disciplines. Courses devoted to parallel programming and writing of parallel CSE applications will benefit from this book.

Arvustused

"It succeeds perfectly and combines for the first time both Charm++ and significant application development in a single volume. It will provide a solid foundation for anyone who is considering using the most recent tools for developing applications for future ExascaIe platforms. I highly recommend this timely book for scientists and engineers." -Horst Simon, Lawrence Berkeley National Laboratory and University of California, Berkeley

List of Figures xiii
List of Tables xxi
Series in Computational Physics xxiii
Foreword xxv
Preface xxvii
About the Editors xxxiii
Contributors xxxv
1 The Charm++ Programming Model 1(16)
Laxmikant V. Kale
Gengbin Zheng
1.1 Design Philosophy
2(1)
1.2 Object-Based Programming Model
3(5)
1.3 Capabilities of the Adaptive Runtime System
8(2)
1.4 Extensions to the Basic Model
10(3)
1.5 Charm++ Ecosystem
13(1)
1.6 Other Languages in the Charm++ Family
14(1)
1.7 Historical Notes
15(1)
1.8 Conclusion
16(1)
2 Designing Charm++ Programs 17(18)
Laxmikant V. Kale
2.1 Simple Stencil: Using Over-Decomposition and Selecting Grain-size
17(6)
2.1.1 Grainsize Decisions
18(3)
2.1.2 Multicore Nodes
21(1)
2.1.3 Migrating Chares, Load Balancing and Fault Tolerance
21(2)
2.2 Multi-Physics Modules Using Multiple Chare Arrays
23(4)
2.2.1 LeanMD
24(3)
2.3 SAMR: Chare Arrays with Dynamic Insertion and Flexible Indices
27(2)
2.4 Combinatorial Search: Task Parallelism
29(1)
2.5 Other Features and Design Considerations
30(1)
2.6 Utility of Charm++ for Future Applications
31(1)
2.7 Summary
32(3)
3 Tools for Debugging and Performance Analysis 35(26)
Filippo Gioachin
Chee Wai Lee
Jonathan Lifflander
Yanhua Sun
Laxmikant V. Kale
3.1 Introduction
36(1)
3.2 Scalable Debugging with CharmDebug
36(11)
3.2.1 Accessing User Information
37(3)
3.2.2 Debugging Problems at Large Scale
40(6)
3.2.3 Summary
46(1)
3.3 Performance Visualization and Analysis via Projections
47(13)
3.3.1 A Simple Projections Primer
49(2)
3.3.2 Features of Projections via Use Cases
51(7)
3.3.3 Advanced Features for Scalable Performance Analysis
58(1)
3.3.4 Summary
59(1)
3.4 Conclusions
60(1)
4 Scalable Molecular Dynamics with NAMD 61(18)
James C. Phillips
Klaus Schulten
Abhinav Bhatele
Chao Mei
Yanhua Sun
Eric J. Bohm
Laxmikant V. Kale
4.1 Introduction
61(1)
4.2 Need for Biomolecular Simulations
62(1)
4.3 Parallel Molecular Dynamics
63(1)
4.4 NAMD's Parallel Design
64(3)
4.4.1 Force Calculations
65(1)
4.4.2 Load Balancing
66(1)
4.5 Enabling Large Simulations
67(5)
4.5.1 Hierarchical Load Balancing
67(1)
4.5.2 SMP Optimizations
68(2)
4.5.3 Optimizing Fine-Grained Communication in NAMD
70(1)
4.5.4 Parallel Input/Output
71(1)
4.6 Scaling Performance
72(3)
4.7 Simulations Enabled by NAMD
75(1)
4.8 Summary
76(3)
5 OpenAtom: Ab initio Molecular Dynamics for Petascale Plat forms 79(26)
Glenn J. Martyna
Eric J. Bohm
Ramprasad Venkataraman
Laxmikant V. Kale
Abhinav Bhatele
5.1 Introduction
80(1)
5.2 Car-Parrinello Molecular Dynamics
81(6)
5.2.1 Density Functional Theory, KS Density Functional Theory and the Local Density Approximation
82(2)
5.2.2 DFT Computations within Basis Sets
84(1)
5.2.3 Molecular Dynamics
84(1)
5.2.4 Ab initio Molecular Dynamics and CPAIMD
84(1)
5.2.5 Path Integrals
85(1)
5.2.6 Parallel Tempering
86(1)
5.3 Parallel Application Design
87(8)
5.3.1 Modular Design and Benefits
87(2)
5.3.2 Parallel Driver
89(4)
5.3.3 Topology Aware Mapping
93(2)
5.4 Charm++ Feature Development
95(2)
5.5 Performance
97(1)
5.6 Impact on Science and Technology
98(5)
5.6.1 Carbon Based Materials for Photovoltaic Applications
98(4)
5.6.2 Metal Insulator Transitions for Novel Devices
102(1)
5.7 Future Work
103(2)
6 N-body Simulations with ChaNGa 105(32)
Thomas R. Quinn
Pritish Jetley
Lazmikant V. Kale
Filippo Gioachin
6.1 Introduction
106(1)
6.2 Code Design
107(8)
6.2.1 Domain Decomposition and Load Balancing
108(2)
6.2.2 Tree Building
110(2)
6.2.3 Tree Walking
112(1)
6.2.4 Force Softening
113(1)
6.2.5 Periodic Boundary Conditions
114(1)
6.2.6 Neighbor Finding
114(1)
6.2.7 Multi-Stepping
115(1)
6.3 Accuracy Tests
115(6)
6.3.1 Force Errors
115(1)
6.3.2 Cosmology Tests
116(5)
6.4 Performance
121(12)
6.4.1 Domain Decomposition and Tree Build Performance
121(2)
6.4.2 Single-Stepping Performance
123(2)
6.4.3 Multi-Stepping Performance
125(1)
6.4.4 ChaNGa on GPUs
125(8)
6.5 Conclusions and Future Work
133(4)
7 Remote Visualization of Cosmological Data Using Salsa 137(12)
Orion Sky Lawlor
Thomas R. Quinn
7.1 Introduction
137(1)
7.2 Salsa Client/Server Rendering Architecture
138(6)
7.2.1 Client Server Communication Styles
139(2)
7.2.2 Image Compression in Salsa
141(1)
7.2.3 GPU Particle Rendering on the Server
142(2)
7.3 Remote Visualization User Interface
144(1)
7.4 Example Use: Galaxy Clusters
145(4)
8 Improving Scalability of BRAMS: a Regional Weather Forecast Model 149(38)
Eduardo R. Rodrigues
Celso L. Mendes
Jairo Panetta
8.1 Introduction
150(1)
8.2 Load Balancing Strategies for Weather Models
151(2)
8.3 The BRAMS Weather Model
153(1)
8.4 Load Balancing Approach
154(4)
8.4.1 Adaptations to AMPI
155(2)
8.4.2 Balancing Algorithms Employed
157(1)
8.5 New Load Balancer
158(4)
8.6 Fully Distributed Strategies
162(4)
8.6.1 Hilbert Curve-Based Load Balancer
162(2)
8.6.2 Diffusion-Based Load Balancer
164(2)
8.7 Experimental Results
166(16)
8.7.1 First Set of Experiments: Privatization Strategy
166(1)
8.7.2 Second Set of Experiments: Virtualization Effects
167(4)
8.7.3 Third Set of Experiments: Centralized Load Balancers
171(3)
8.7.4 Fourth Set of Experiments: Distributed Load Balancers
174(8)
8.8 Final Remarks
182(5)
9 Crack Propagation Analysis with Automatic Load Balancing 187(24)
Orion Sky Lawlor
M. Scot Breitenfeld
Philippe H. Geubelle
Gengbin Zheng
9.1 Introduction
187(6)
9.1.1 ParFUM Framework
188(2)
9.1.2 Implementation of the ParFUM Framework
190(3)
9.2 Load Balancing Finite Element Codes in Charm++
193(6)
9.2.1 Runtime Support for Thread Migration
193(1)
9.2.2 Comparison to Prior Work
194(1)
9.2.3 Automatic Load Balancing for FEM
195(1)
9.2.4 Load Balancing Strategies
196(1)
9.2.5 Agile Load Balancing
197(2)
9.3 Cohesive and Elasto-plastic Finite Element Model of Fracture
199(11)
9.3.1 Case Study 1: Elasto-Plastic Wave Propagation
202(4)
9.3.2 Case Study 2: Dynamic Fracture
206(4)
9.4 Conclusions
210(1)
10 Contagion Diffusion with EpiSimdemics 211(36)
Keith R. Bisset
Ashwin M. Aji
Tariq Kamal
Jae-Seung Yeom
Madhav V. Marathe
Eric J. Bohm
Abhishek Gupta
10.1 Introduction
212(3)
10.2 Problem Description
215(3)
10.2.1 Formalization
215(2)
10.2.2 Application to Computational Epidemiology
217(1)
10.3 EpiSimdemics Design
218(6)
10.3.1 The Disease Model
220(1)
10.3.2 Modeling Behavior of Individual Agents
220(1)
10.3.3 Intervention and Behavior Modification
221(2)
10.3.4 Social Network Representation
223(1)
10.4 EpiSimdemics Algorithm
224(3)
10.5 Charm++ Implementation
227(5)
10.5.1 Designing the Chares
228(2)
10.5.2 The EpiSimdemics Algorithm
230(2)
10.5.3 Charm++ Features
232(1)
10.6 Performance of EpiSimdemics
232(9)
10.6.1 Experimental Setup
233(1)
10.6.2 Performance Characteristics
233(1)
10.6.3 Effects of Synchronization
234(1)
10.6.4 Effects of Strong Scaling
235(1)
10.6.5 Effects of Weak Scaling
236(1)
10.6.6 Effect of Load Balancing
237(4)
10.7 Representative Study
241(6)
Bibliography 247(24)
Index 271
Laxmikant V. Kale, Abhinav Bhatele