Jeannot
John Wiley & Sons
e druk, 2014
9781118712054
High–Performance Computing on Complex Environments
Specificaties
Gebonden, 512 blz.
|
Engels
John Wiley & Sons |
e druk, 2014
ISBN13: 9781118712054
Rubricering
Onderdeel van serie
Wiley Series on Parallel and Distributed Computing
Verwachte levertijd ongeveer 9 werkdagen
Samenvatting
With recent changes in multicore and general–purpose computing on graphics processing units, the way parallel computers are used and programmed has drastically changed. It is important to provide a comprehensive study on how to use such machines written by specialists of the domain. The book provides recent research results in high–performance computing on complex environments, information on how to efficiently exploit heterogeneous and hierarchical architectures and distributed systems, detailed studies on the impact of applying heterogeneous computing practices to real problems, and applications varying from remote sensing to tomography. The content spans topics such as Numerical Analysis for Heterogeneous and Multicore Systems; Optimization of Communication for High Performance Heterogeneous and Hierarchical Platforms; Efficient Exploitation of Heterogeneous Architectures, Hybrid CPU+GPU, and Distributed Systems; Energy Awareness in High–Performance Computing; and Applications of Heterogeneous High–Performance Computing.
Covers cutting–edge research in HPC on complex environments, following an international collaboration of members of the ComplexHPC
Explains how to efficiently exploit heterogeneous and hierarchical architectures and distributed systems
Twenty–three chapters and over 100 illustrations cover domains such as numerical analysis, communication and storage, applications, GPUs and accelerators, and energy efficiency
Specificaties
Inhoudsopgave
<p>Contributors xxiii</p>
<p>Preface xxvii</p>
<p>PART I INTRODUCTION 1</p>
<p>1. Summary of the Open European Network for High–Performance Computing in Complex Environments 3<br /> Emmanuel Jeannot and Julius Zilinskas</p>
<p>1.1 Introduction and Vision 4</p>
<p>1.2 Scientific Organization 6</p>
<p>1.3 Activities of the Project 6</p>
<p>1.4 Main Outcomes of the Action 7</p>
<p>1.5 Contents of the Book 8</p>
<p>PART II NUMERICAL ANALYSIS FOR HETEROGENEOUS AND MULTICORE SYSTEMS 11</p>
<p>2. On the Impact of the Heterogeneous Multicore and Many–Core Platforms on Iterative Solution Methods and Preconditioning Techniques 13<br /> Dimitar Lukarski and Maya Neytcheva</p>
<p>2.1 Introduction 14</p>
<p>2.2 General Description of Iterative Methods and Preconditioning 16</p>
<p>2.3 Preconditioning Techniques 20</p>
<p>2.4 Defect–Correction Technique 21</p>
<p>2.5 Multigrid Method 22</p>
<p>2.6 Parallelization of Iterative Methods 22</p>
<p>2.7 Heterogeneous Systems 23</p>
<p>2.8 Maintenance and Portability 29</p>
<p>2.9 Conclusion 30</p>
<p>3. Efficient Numerical Solution of 2D Diffusion Equation on Multicore Computers 33<br /> Matjaz Depolli, Gregor Kosec, and Roman Trobec</p>
<p>3.1 Introduction 34</p>
<p>3.2 Test Case 35</p>
<p>3.3 Parallel Implementation 39</p>
<p>3.4 Results 41</p>
<p>3.5 Discussion 45</p>
<p>3.6 Conclusion 47</p>
<p>4. Parallel Algorithms for Parabolic Problems on Graphs in Neuroscience 51<br /> Natalija Tumanova and Raimondas Ciegis</p>
<p>4.1 Introduction 51</p>
<p>4.2 Formulation of the Discrete Model 53</p>
<p>4.3 Parallel Algorithms 59</p>
<p>4.4 Computational Results 63</p>
<p>4.5 Conclusions 69</p>
<p>PART III COMMUNICATION AND STORAGE CONSIDERATIONS IN HIGH–PERFORMANCE COMPUTING 73</p>
<p>5. An Overview of Topology Mapping Algorithms and Techniques in High–Performance Computing 75<br /> Torsten Hoefler, Emmanuel Jeannot, and Guillaume Mercier</p>
<p>5.1 Introduction 76</p>
<p>5.2 General Overview 76</p>
<p>5.3 Formalization of the Problem 79</p>
<p>5.4 Algorithmic Strategies for Topology Mapping 81</p>
<p>5.5 Mapping Enforcement Techniques 82</p>
<p>5.6 Survey of Solutions 85</p>
<p>5.7 Conclusion and Open Problems 89</p>
<p>6. Optimization of Collective Communication for Heterogeneous HPC Platforms 95<br /> Kiril Dichev and Alexey Lastovetsky</p>
<p>6.1 Introduction 95</p>
<p>6.2 Overview of Optimized Collectives and Topology–Aware Collectives 97</p>
<p>6.3 Optimizations of Collectives on Homogeneous Clusters 98</p>
<p>6.4 Heterogeneous Networks 99</p>
<p>6.5 Topology– and Performance–Aware Collectives 100</p>
<p>6.6 Topology as Input 101</p>
<p>6.7 Performance as Input 102</p>
<p>6.8 Non–MPI Collective Algorithms for Heterogeneous Networks 106</p>
<p>6.9 Conclusion 111</p>
<p>7. Effective Data Access Patterns on Massively Parallel Processors 115<br /> Gabriele Capannini, Ranieri Baraglia, Fabrizio Silvestri, and Franco Maria Nardini</p>
<p>7.1 Introduction 115</p>
<p>7.2 Architectural Details 116</p>
<p>7.3 K–Model 117</p>
<p>7.4 Parallel Prefix Sum 120</p>
<p>7.5 Bitonic Sorting Networks 126</p>
<p>7.6 Final Remarks 132</p>
<p>8. Scalable Storage I/O Software for Blue Gene Architectures 135<br /> Florin Isaila, Javier Garcia, and Jesús Carretero</p>
<p>8.1 Introduction 135</p>
<p>8.2 Blue Gene System Overview 136</p>
<p>8.3 Design and Implementation 138</p>
<p>8.4 Conclusions and Future Work 142</p>
<p>PART IV EFFICIENT EXPLOITATION OF HETEROGENEOUS ARCHITECTURES 145</p>
<p>9. Fair Resource Sharing for Dynamic Scheduling of Workflows on Heterogeneous Systems 147<br /> Hamid Arabnejad, Jorge G. Barbosa, and Frédéric Suter</p>
<p>9.1 Introduction 148</p>
<p>9.2 Concurrent Workflow Scheduling 153</p>
<p>9.3 Experimental Results and Discussion 160</p>
<p>9.4 Conclusions 165</p>
<p>10. Systematic Mapping of Reed Solomon Erasure Codes on Heterogeneous Multicore Architectures 169<br /> Roman Wyrzykowski, Marcin Wozniak, and Lukasz Kuczynski</p>
<p>10.1 Introduction 169</p>
<p>10.2 Related Works 171</p>
<p>10.3 Reed Solomon Codes and Linear Algebra Algorithms 172</p>
<p>10.4 Mapping Reed Solomon Codes on Cell/B.E. Architecture 173</p>
<p>10.5 Mapping Reed Solomon Codes on Multicore GPU Architectures 178</p>
<p>10.6 Methods of Increasing the Algorithm Performance on GPUs 181</p>
<p>10.7 GPU Performance Evaluation 185</p>
<p>10.8 Conclusions and Future Works 190</p>
<p>11. Heterogeneous Parallel Computing Platforms and Tools for Compute–Intensive Algorithms: A Case Study 193<br /> Daniele D′Agostino, Andrea Clematis, and Emanuele Danovaro</p>
<p>11.1 Introduction 194</p>
<p>11.2 A Low–Cost Heterogeneous Computing Environment 196</p>
<p>11.3 First Case Study: The N–Body Problem 200</p>
<p>11.4 Second Case Study: The Convolution Algorithm 206</p>
<p>11.5 Conclusions 211</p>
<p>12. Efficient Application of Hybrid Parallelism in Electromagnetism Problems 215<br /> Alejandro Alvarez–Melcon, Fernando D. Quesada, Domingo Gimenez, Carlos Pérez–Alcaraz, Jose–Gines Picon, and Tomas Ramírez</p>
<p>12.1 Introduction 215</p>
<p>12.2 Computation of Green s functions in Hybrid Systems 216</p>
<p>12.3 Parallelization in Numa Systems of a Volume Integral Equation Technique 222</p>
<p>12.4 Autotuning Parallel Codes 226</p>
<p>12.5 Conclusions and Future Research 230</p>
<p>PART V CPU + GPU COPROCESSING 235</p>
<p>13. Design and Optimization of Scientific Applications for Highly Heterogeneous and Hierarchical HPC Platforms Using Functional Computation Performance Models 237<br /> David Clarke, Aleksandar Ilic, Alexey Lastovetsky, Vladimir Rychkov, Leonel Sousa, and Ziming Zhong</p>
<p>13.1 Introduction 238</p>
<p>13.2 Related Work 241</p>
<p>13.3 Data Partitioning Based on Functional Performance Model 243</p>
<p>13.4 Example Application: Heterogeneous Parallel Matrix Multiplication 245</p>
<p>13.5 Performance Measurement on CPUs/GPUs System 247</p>
<p>13.6 Functional Performance Models of Multiple Cores and GPUs 248</p>
<p>13.7 FPM–Based Data Partitioning on CPUs/GPUs System 250</p>
<p>13.8 Efficient Building of Functional Performance Models 251</p>
<p>13.9 FPM–Based Data Partitioning on Hierarchical Platforms 253</p>
<p>13.10 Conclusion 257</p>
<p>14. Efficient Multilevel Load Balancing on Heterogeneous CPU + GPU Systems 261<br /> Aleksandar Ilic and Leonel Sousa</p>
<p>14.1 Introduction: Heterogeneous CPU + GPU Systems 262</p>
<p>14.2 Background and Related Work 265</p>
<p>14.3 Load Balancing Algorithms for Heterogeneous CPU + GPU Systems 269</p>
<p>14.4 Experimental Results 275</p>
<p>14.5 Conclusions 279</p>
<p>15. The All–Pair Shortest–Path Problem in Shared–Memory Heterogeneous Systems 283<br /> Hector Ortega–Arranz, Yuri Torres, Diego R. Llanos, and Arturo Gonzalez–Escribano</p>
<p>15.1 Introduction 283</p>
<p>15.2 Algorithmic Overview 285</p>
<p>15.3 CUDA Overview 287</p>
<p>15.4 Heterogeneous Systems and Load Balancing 288</p>
<p>15.5 Parallel Solutions to The APSP 289</p>
<p>15.6 Experimental Setup 291</p>
<p>15.7 Experimental Results 293</p>
<p>15.8 Conclusions 297</p>
<p>PART VI EFFICIENT EXPLOITATION OF DISTRIBUTED SYSTEMS 301</p>
<p>16. Resource Management for HPC on the Cloud 303<br /> Marc E. Frincu and Dana Petcu</p>
<p>16.1 Introduction 303</p>
<p>16.2 On the Type of Applications for HPC and HPC2 305</p>
<p>16.3 HPC on the Cloud 306</p>
<p>16.4 Scheduling Algorithms for HPC2 311</p>
<p>16.5 Toward an Autonomous Scheduling Framework 312</p>
<p>16.6 Conclusions 319</p>
<p>17. Resource Discovery in Large–Scale Grid Systems 323<br /> Konstantinos Karaoglanoglou and Helen Karatza</p>
<p>17.1 Introduction and Background 323</p>
<p>17.2 The Semantic Communities Approach 325</p>
<p>17.3 The P2P Approach 329</p>
<p>17.4 The Grid–Routing Transferring Approach 333</p>
<p>17.5 Conclusions 337</p>
<p>PART VII ENERGY AWARENESS IN HIGH–PERFORMANCE COMPUTING 341</p>
<p>18. Energy–Aware Approaches for HPC Systems 343<br /> Robert Basmadjian, Georges Da Costa, Ghislain Landry Tsafack Chetsa, Laurent Lefevre, Ariel Oleksiak, and Jean–Marc Pierson</p>
<p>18.1 Introduction 344</p>
<p>18.2 Power Consumption of Servers 345</p>
<p>18.3 Classification and Energy Profiles of HPC Applications 354</p>
<p>18.4 Policies and Leverages 359</p>
<p>18.5 Conclusion 360</p>
<p>19. Strategies for Increased Energy Awareness in Cloud Federations 365<br /> Gabor Kecskemeti, AttilaKertesz, Attila Cs. Marosi, and Zsolt Nemeth</p>
<p>19.1 Introduction 365</p>
<p>19.2 Related Work 367</p>
<p>19.3 Scenarios 369</p>
<p>19.4 Energy–Aware Cloud Federations 374</p>
<p>19.5 Conclusions 379</p>
<p>20. Enabling Network Security in HPC Systems Using Heterogeneous CMPs 383<br /> Ozcan Ozturk and Suleyman Tosun</p>
<p>20.1 Introduction 384</p>
<p>20.2 Related Work 386</p>
<p>20.3 Overview of Our Approach 387</p>
<p>20.4 Heterogeneous CMP Design for Network Security Processors 390</p>
<p>20.5 Experimental Evaluation 394</p>
<p>20.6 Concluding Remarks 397</p>
<p>PART VIII APPLICATIONS OF HETEROGENEOUS HIGH–PERFORMANCE COMPUTING 401</p>
<p>21. Toward a High–Performance Distributed CBIR System for Hyperspectral Remote Sensing Data: A Case Study in Jungle Computing 403<br /> Timo van Kessel, NielsDrost, Jason Maassen, Henri E. Bal, Frank J. Seinstra, and Antonio J. Plaza</p>
<p>21.1 Introduction 404</p>
<p>21.2 CBIR For Hyperspectral Imaging Data 407</p>
<p>21.3 Jungle Computing 410</p>
<p>21.4 IBIS and Constellation 412</p>
<p>21.5 System Design and Implementation 415</p>
<p>21.6 Evaluation 420</p>
<p>21.7 Conclusions 426</p>
<p>22. Taking Advantage of Heterogeneous Platforms in Image and Video Processing 429<br /> Sidi A. Mahmoudi, Erencan Ozkan, Pierre Manneback, and Suleyman Tosun</p>
<p>22.1 Introduction 430</p>
<p>22.2 Related Work 431</p>
<p>22.3 Parallel Image Processing on GPU 433</p>
<p>22.4 Image Processing on Heterogeneous Architectures 437</p>
<p>22.5 Video Processing on GPU 438</p>
<p>22.6 Experimental Results 444</p>
<p>22.7 Conclusion 447</p>
<p>23. Real–Time Tomographic Reconstruction Through CPU + GPU Coprocessing 451<br /> Jose Ignacio Agulleiro, Francisco Vazquez, Ester M. Garzon, and Jose J. Fernandez</p>
<p>23.1 Introduction 452</p>
<p>23.2 Tomographic Reconstruction 453</p>
<p>23.3 Optimization of Tomographic Reconstruction for CPUs and for GPUs 455</p>
<p>23.4 Hybrid CPU + GPU Tomographic Reconstruction 457</p>
<p>23.5 Results 459</p>
<p>23.6 Discussion and Conclusion 461</p>
<p>Acknowledgments 463</p>
<p>References 463</p>
<p>Index 467</p>
<p>Preface xxvii</p>
<p>PART I INTRODUCTION 1</p>
<p>1. Summary of the Open European Network for High–Performance Computing in Complex Environments 3<br /> Emmanuel Jeannot and Julius Zilinskas</p>
<p>1.1 Introduction and Vision 4</p>
<p>1.2 Scientific Organization 6</p>
<p>1.3 Activities of the Project 6</p>
<p>1.4 Main Outcomes of the Action 7</p>
<p>1.5 Contents of the Book 8</p>
<p>PART II NUMERICAL ANALYSIS FOR HETEROGENEOUS AND MULTICORE SYSTEMS 11</p>
<p>2. On the Impact of the Heterogeneous Multicore and Many–Core Platforms on Iterative Solution Methods and Preconditioning Techniques 13<br /> Dimitar Lukarski and Maya Neytcheva</p>
<p>2.1 Introduction 14</p>
<p>2.2 General Description of Iterative Methods and Preconditioning 16</p>
<p>2.3 Preconditioning Techniques 20</p>
<p>2.4 Defect–Correction Technique 21</p>
<p>2.5 Multigrid Method 22</p>
<p>2.6 Parallelization of Iterative Methods 22</p>
<p>2.7 Heterogeneous Systems 23</p>
<p>2.8 Maintenance and Portability 29</p>
<p>2.9 Conclusion 30</p>
<p>3. Efficient Numerical Solution of 2D Diffusion Equation on Multicore Computers 33<br /> Matjaz Depolli, Gregor Kosec, and Roman Trobec</p>
<p>3.1 Introduction 34</p>
<p>3.2 Test Case 35</p>
<p>3.3 Parallel Implementation 39</p>
<p>3.4 Results 41</p>
<p>3.5 Discussion 45</p>
<p>3.6 Conclusion 47</p>
<p>4. Parallel Algorithms for Parabolic Problems on Graphs in Neuroscience 51<br /> Natalija Tumanova and Raimondas Ciegis</p>
<p>4.1 Introduction 51</p>
<p>4.2 Formulation of the Discrete Model 53</p>
<p>4.3 Parallel Algorithms 59</p>
<p>4.4 Computational Results 63</p>
<p>4.5 Conclusions 69</p>
<p>PART III COMMUNICATION AND STORAGE CONSIDERATIONS IN HIGH–PERFORMANCE COMPUTING 73</p>
<p>5. An Overview of Topology Mapping Algorithms and Techniques in High–Performance Computing 75<br /> Torsten Hoefler, Emmanuel Jeannot, and Guillaume Mercier</p>
<p>5.1 Introduction 76</p>
<p>5.2 General Overview 76</p>
<p>5.3 Formalization of the Problem 79</p>
<p>5.4 Algorithmic Strategies for Topology Mapping 81</p>
<p>5.5 Mapping Enforcement Techniques 82</p>
<p>5.6 Survey of Solutions 85</p>
<p>5.7 Conclusion and Open Problems 89</p>
<p>6. Optimization of Collective Communication for Heterogeneous HPC Platforms 95<br /> Kiril Dichev and Alexey Lastovetsky</p>
<p>6.1 Introduction 95</p>
<p>6.2 Overview of Optimized Collectives and Topology–Aware Collectives 97</p>
<p>6.3 Optimizations of Collectives on Homogeneous Clusters 98</p>
<p>6.4 Heterogeneous Networks 99</p>
<p>6.5 Topology– and Performance–Aware Collectives 100</p>
<p>6.6 Topology as Input 101</p>
<p>6.7 Performance as Input 102</p>
<p>6.8 Non–MPI Collective Algorithms for Heterogeneous Networks 106</p>
<p>6.9 Conclusion 111</p>
<p>7. Effective Data Access Patterns on Massively Parallel Processors 115<br /> Gabriele Capannini, Ranieri Baraglia, Fabrizio Silvestri, and Franco Maria Nardini</p>
<p>7.1 Introduction 115</p>
<p>7.2 Architectural Details 116</p>
<p>7.3 K–Model 117</p>
<p>7.4 Parallel Prefix Sum 120</p>
<p>7.5 Bitonic Sorting Networks 126</p>
<p>7.6 Final Remarks 132</p>
<p>8. Scalable Storage I/O Software for Blue Gene Architectures 135<br /> Florin Isaila, Javier Garcia, and Jesús Carretero</p>
<p>8.1 Introduction 135</p>
<p>8.2 Blue Gene System Overview 136</p>
<p>8.3 Design and Implementation 138</p>
<p>8.4 Conclusions and Future Work 142</p>
<p>PART IV EFFICIENT EXPLOITATION OF HETEROGENEOUS ARCHITECTURES 145</p>
<p>9. Fair Resource Sharing for Dynamic Scheduling of Workflows on Heterogeneous Systems 147<br /> Hamid Arabnejad, Jorge G. Barbosa, and Frédéric Suter</p>
<p>9.1 Introduction 148</p>
<p>9.2 Concurrent Workflow Scheduling 153</p>
<p>9.3 Experimental Results and Discussion 160</p>
<p>9.4 Conclusions 165</p>
<p>10. Systematic Mapping of Reed Solomon Erasure Codes on Heterogeneous Multicore Architectures 169<br /> Roman Wyrzykowski, Marcin Wozniak, and Lukasz Kuczynski</p>
<p>10.1 Introduction 169</p>
<p>10.2 Related Works 171</p>
<p>10.3 Reed Solomon Codes and Linear Algebra Algorithms 172</p>
<p>10.4 Mapping Reed Solomon Codes on Cell/B.E. Architecture 173</p>
<p>10.5 Mapping Reed Solomon Codes on Multicore GPU Architectures 178</p>
<p>10.6 Methods of Increasing the Algorithm Performance on GPUs 181</p>
<p>10.7 GPU Performance Evaluation 185</p>
<p>10.8 Conclusions and Future Works 190</p>
<p>11. Heterogeneous Parallel Computing Platforms and Tools for Compute–Intensive Algorithms: A Case Study 193<br /> Daniele D′Agostino, Andrea Clematis, and Emanuele Danovaro</p>
<p>11.1 Introduction 194</p>
<p>11.2 A Low–Cost Heterogeneous Computing Environment 196</p>
<p>11.3 First Case Study: The N–Body Problem 200</p>
<p>11.4 Second Case Study: The Convolution Algorithm 206</p>
<p>11.5 Conclusions 211</p>
<p>12. Efficient Application of Hybrid Parallelism in Electromagnetism Problems 215<br /> Alejandro Alvarez–Melcon, Fernando D. Quesada, Domingo Gimenez, Carlos Pérez–Alcaraz, Jose–Gines Picon, and Tomas Ramírez</p>
<p>12.1 Introduction 215</p>
<p>12.2 Computation of Green s functions in Hybrid Systems 216</p>
<p>12.3 Parallelization in Numa Systems of a Volume Integral Equation Technique 222</p>
<p>12.4 Autotuning Parallel Codes 226</p>
<p>12.5 Conclusions and Future Research 230</p>
<p>PART V CPU + GPU COPROCESSING 235</p>
<p>13. Design and Optimization of Scientific Applications for Highly Heterogeneous and Hierarchical HPC Platforms Using Functional Computation Performance Models 237<br /> David Clarke, Aleksandar Ilic, Alexey Lastovetsky, Vladimir Rychkov, Leonel Sousa, and Ziming Zhong</p>
<p>13.1 Introduction 238</p>
<p>13.2 Related Work 241</p>
<p>13.3 Data Partitioning Based on Functional Performance Model 243</p>
<p>13.4 Example Application: Heterogeneous Parallel Matrix Multiplication 245</p>
<p>13.5 Performance Measurement on CPUs/GPUs System 247</p>
<p>13.6 Functional Performance Models of Multiple Cores and GPUs 248</p>
<p>13.7 FPM–Based Data Partitioning on CPUs/GPUs System 250</p>
<p>13.8 Efficient Building of Functional Performance Models 251</p>
<p>13.9 FPM–Based Data Partitioning on Hierarchical Platforms 253</p>
<p>13.10 Conclusion 257</p>
<p>14. Efficient Multilevel Load Balancing on Heterogeneous CPU + GPU Systems 261<br /> Aleksandar Ilic and Leonel Sousa</p>
<p>14.1 Introduction: Heterogeneous CPU + GPU Systems 262</p>
<p>14.2 Background and Related Work 265</p>
<p>14.3 Load Balancing Algorithms for Heterogeneous CPU + GPU Systems 269</p>
<p>14.4 Experimental Results 275</p>
<p>14.5 Conclusions 279</p>
<p>15. The All–Pair Shortest–Path Problem in Shared–Memory Heterogeneous Systems 283<br /> Hector Ortega–Arranz, Yuri Torres, Diego R. Llanos, and Arturo Gonzalez–Escribano</p>
<p>15.1 Introduction 283</p>
<p>15.2 Algorithmic Overview 285</p>
<p>15.3 CUDA Overview 287</p>
<p>15.4 Heterogeneous Systems and Load Balancing 288</p>
<p>15.5 Parallel Solutions to The APSP 289</p>
<p>15.6 Experimental Setup 291</p>
<p>15.7 Experimental Results 293</p>
<p>15.8 Conclusions 297</p>
<p>PART VI EFFICIENT EXPLOITATION OF DISTRIBUTED SYSTEMS 301</p>
<p>16. Resource Management for HPC on the Cloud 303<br /> Marc E. Frincu and Dana Petcu</p>
<p>16.1 Introduction 303</p>
<p>16.2 On the Type of Applications for HPC and HPC2 305</p>
<p>16.3 HPC on the Cloud 306</p>
<p>16.4 Scheduling Algorithms for HPC2 311</p>
<p>16.5 Toward an Autonomous Scheduling Framework 312</p>
<p>16.6 Conclusions 319</p>
<p>17. Resource Discovery in Large–Scale Grid Systems 323<br /> Konstantinos Karaoglanoglou and Helen Karatza</p>
<p>17.1 Introduction and Background 323</p>
<p>17.2 The Semantic Communities Approach 325</p>
<p>17.3 The P2P Approach 329</p>
<p>17.4 The Grid–Routing Transferring Approach 333</p>
<p>17.5 Conclusions 337</p>
<p>PART VII ENERGY AWARENESS IN HIGH–PERFORMANCE COMPUTING 341</p>
<p>18. Energy–Aware Approaches for HPC Systems 343<br /> Robert Basmadjian, Georges Da Costa, Ghislain Landry Tsafack Chetsa, Laurent Lefevre, Ariel Oleksiak, and Jean–Marc Pierson</p>
<p>18.1 Introduction 344</p>
<p>18.2 Power Consumption of Servers 345</p>
<p>18.3 Classification and Energy Profiles of HPC Applications 354</p>
<p>18.4 Policies and Leverages 359</p>
<p>18.5 Conclusion 360</p>
<p>19. Strategies for Increased Energy Awareness in Cloud Federations 365<br /> Gabor Kecskemeti, AttilaKertesz, Attila Cs. Marosi, and Zsolt Nemeth</p>
<p>19.1 Introduction 365</p>
<p>19.2 Related Work 367</p>
<p>19.3 Scenarios 369</p>
<p>19.4 Energy–Aware Cloud Federations 374</p>
<p>19.5 Conclusions 379</p>
<p>20. Enabling Network Security in HPC Systems Using Heterogeneous CMPs 383<br /> Ozcan Ozturk and Suleyman Tosun</p>
<p>20.1 Introduction 384</p>
<p>20.2 Related Work 386</p>
<p>20.3 Overview of Our Approach 387</p>
<p>20.4 Heterogeneous CMP Design for Network Security Processors 390</p>
<p>20.5 Experimental Evaluation 394</p>
<p>20.6 Concluding Remarks 397</p>
<p>PART VIII APPLICATIONS OF HETEROGENEOUS HIGH–PERFORMANCE COMPUTING 401</p>
<p>21. Toward a High–Performance Distributed CBIR System for Hyperspectral Remote Sensing Data: A Case Study in Jungle Computing 403<br /> Timo van Kessel, NielsDrost, Jason Maassen, Henri E. Bal, Frank J. Seinstra, and Antonio J. Plaza</p>
<p>21.1 Introduction 404</p>
<p>21.2 CBIR For Hyperspectral Imaging Data 407</p>
<p>21.3 Jungle Computing 410</p>
<p>21.4 IBIS and Constellation 412</p>
<p>21.5 System Design and Implementation 415</p>
<p>21.6 Evaluation 420</p>
<p>21.7 Conclusions 426</p>
<p>22. Taking Advantage of Heterogeneous Platforms in Image and Video Processing 429<br /> Sidi A. Mahmoudi, Erencan Ozkan, Pierre Manneback, and Suleyman Tosun</p>
<p>22.1 Introduction 430</p>
<p>22.2 Related Work 431</p>
<p>22.3 Parallel Image Processing on GPU 433</p>
<p>22.4 Image Processing on Heterogeneous Architectures 437</p>
<p>22.5 Video Processing on GPU 438</p>
<p>22.6 Experimental Results 444</p>
<p>22.7 Conclusion 447</p>
<p>23. Real–Time Tomographic Reconstruction Through CPU + GPU Coprocessing 451<br /> Jose Ignacio Agulleiro, Francisco Vazquez, Ester M. Garzon, and Jose J. Fernandez</p>
<p>23.1 Introduction 452</p>
<p>23.2 Tomographic Reconstruction 453</p>
<p>23.3 Optimization of Tomographic Reconstruction for CPUs and for GPUs 455</p>
<p>23.4 Hybrid CPU + GPU Tomographic Reconstruction 457</p>
<p>23.5 Results 459</p>
<p>23.6 Discussion and Conclusion 461</p>
<p>Acknowledgments 463</p>
<p>References 463</p>
<p>Index 467</p>