{"product_id":"energy-efficient-distributed-computing-systems-hardback-9780470908754","title":"Energy-Efficient Distributed Computing Systems (Hardback) 9780470908754","description":"\u003cfont face=\"Georgia\"\u003e\r\n\u003cp\u003e\u003cfont size=\"6\"\u003eEnergy-Efficient Distributed Computing Systems\u003c\/font\u003e\u003cbr\u003e\r\n\r\n\r\n\r\n\r\n\r\n\u003c\/p\u003e\n\u003cp\u003e\u003cfont size=\"4\"\u003eAlbert Y. Zomaya (Edited by), AY Zomaya (Author), Young Choon Lee (Edited by)\u003c\/font\u003e\u003c\/p\u003e\r\n\r\n\u003cp\u003e\u003cfont size=\"3\"\u003e9780470908754, Wiley\u003c\/font\u003e\u003c\/p\u003e\r\n\r\n\u003cp\u003e\u003cfont size=\"3\"\u003eHardback, published 14 September 2012\u003c\/font\u003e\u003c\/p\u003e\r\n\r\n\u003cp\u003e\u003cfont size=\"3\"\u003e856 pages\u003cbr\u003e23.9 x 16.3 x 5.1 cm, 1.27 kg\u003c\/font\u003e\u003c\/p\u003e\r\n\r\n\r\n\r\n\r\n\r\n\u003cp align=\"justify\"\u003e\u003cstrong\u003e\u003cfont size=\"3\"\u003eThe energy consumption issue in distributed computing systems raises various monetary, environmental and system performance concerns. Electricity consumption in the US doubled from 2000 to 2005.  From a financial and environmental standpoint, reducing the consumption of electricity is important, yet these reforms must not lead to performance degradation of the computing systems.  These contradicting constraints create a suite of complex problems that need to be resolved in order to lead to 'greener' distributed computing systems.  This book brings together a group of outstanding researchers that investigate the different facets of green and energy efficient distributed computing.\u003cbr\u003e \u003cbr\u003e   \u003cp\u003eKey features:\u003c\/p\u003e \u003cul\u003e \u003cli\u003eOne of the first books of its kind\u003c\/li\u003e \u003cli\u003eFeatures latest research findings on emerging topics by well-known scientists\u003c\/li\u003e \u003cli\u003eValuable research for grad students, postdocs, and researchers\u003c\/li\u003e \u003cli\u003eResearch will greatly feed into other technologies and application domains\u003c\/li\u003e \u003c\/ul\u003e\u003c\/font\u003e\u003c\/strong\u003e\u003c\/p\u003e\r\n\r\n\u003cp\u003e\u003cfont size=\"3\"\u003ePREFACE xxix  \u003cp\u003eACKNOWLEDGMENTS xxxi\u003c\/p\u003e \u003cp\u003eCONTRIBUTORS xxxiii\u003c\/p\u003e \u003cp\u003e\u003cb\u003e1 POWER ALLOCATION AND TASK SCHEDULING ON MULTIPROCESSOR COMPUTERS WITH ENERGY AND TIME CONSTRAINTS 1\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eKeqin Li\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e1.1 Introduction 1\u003c\/p\u003e \u003cp\u003e1.1.1 Energy Consumption 1\u003c\/p\u003e \u003cp\u003e1.1.2 Power Reduction 2\u003c\/p\u003e \u003cp\u003e1.1.3 Dynamic Power Management 3\u003c\/p\u003e \u003cp\u003e1.1.4 Task Scheduling with Energy and Time Constraints 4\u003c\/p\u003e \u003cp\u003e1.1.5 Chapter Outline 5\u003c\/p\u003e \u003cp\u003e1.2 Preliminaries 5\u003c\/p\u003e \u003cp\u003e1.2.1 Power Consumption Model 5\u003c\/p\u003e \u003cp\u003e1.2.2 Problem Definitions 6\u003c\/p\u003e \u003cp\u003e1.2.3 Task Models 7\u003c\/p\u003e \u003cp\u003e1.2.4 Processor Models 8\u003c\/p\u003e \u003cp\u003e1.2.5 Scheduling Models 9\u003c\/p\u003e \u003cp\u003e1.2.6 Problem Decomposition 9\u003c\/p\u003e \u003cp\u003e1.2.7 Types of Algorithms 10\u003c\/p\u003e \u003cp\u003e1.3 Problem Analysis 10\u003c\/p\u003e \u003cp\u003e1.3.1 Schedule Length Minimization 10\u003c\/p\u003e \u003cp\u003e1.3.1.1 Uniprocessor computers 10\u003c\/p\u003e \u003cp\u003e1.3.1.2 Multiprocessor computers 11\u003c\/p\u003e \u003cp\u003e1.3.2 Energy Consumption Minimization 12\u003c\/p\u003e \u003cp\u003e1.3.2.1 Uniprocessor computers 12\u003c\/p\u003e \u003cp\u003e1.3.2.2 Multiprocessor computers 13\u003c\/p\u003e \u003cp\u003e1.3.3 Strong NP-Hardness 14\u003c\/p\u003e \u003cp\u003e1.3.4 Lower Bounds 14\u003c\/p\u003e \u003cp\u003e1.3.5 Energy-Delay Trade-off 15\u003c\/p\u003e \u003cp\u003e1.4 Pre-Power-Determination Algorithms 16\u003c\/p\u003e \u003cp\u003e1.4.1 Overview 16\u003c\/p\u003e \u003cp\u003e1.4.2 Performance Measures 17\u003c\/p\u003e \u003cp\u003e1.4.3 Equal-Time Algorithms and Analysis 18\u003c\/p\u003e \u003cp\u003e1.4.3.1 Schedule length minimization 18\u003c\/p\u003e \u003cp\u003e1.4.3.2 Energy consumption minimization 19\u003c\/p\u003e \u003cp\u003e1.4.4 Equal-Energy Algorithms and Analysis 19\u003c\/p\u003e \u003cp\u003e1.4.4.1 Schedule length minimization 19\u003c\/p\u003e \u003cp\u003e1.4.4.2 Energy consumption minimization 21\u003c\/p\u003e \u003cp\u003e1.4.5 Equal-Speed Algorithms and Analysis 22\u003c\/p\u003e \u003cp\u003e1.4.5.1 Schedule length minimization 22\u003c\/p\u003e \u003cp\u003e1.4.5.2 Energy consumption minimization 23\u003c\/p\u003e \u003cp\u003e1.4.6 Numerical Data 24\u003c\/p\u003e \u003cp\u003e1.4.7 Simulation Results 25\u003c\/p\u003e \u003cp\u003e1.5 Post-Power-Determination Algorithms 28\u003c\/p\u003e \u003cp\u003e1.5.1 Overview 28\u003c\/p\u003e \u003cp\u003e1.5.2 Analysis of List Scheduling Algorithms 29\u003c\/p\u003e \u003cp\u003e1.5.2.1 Analysis of algorithm LS 29\u003c\/p\u003e \u003cp\u003e1.5.2.2 Analysis of algorithm LRF 30\u003c\/p\u003e \u003cp\u003e1.5.3 Application to Schedule Length Minimization 30\u003c\/p\u003e \u003cp\u003e1.5.4 Application to Energy Consumption Minimization 31\u003c\/p\u003e \u003cp\u003e1.5.5 Numerical Data 32\u003c\/p\u003e \u003cp\u003e1.5.6 Simulation Results 32\u003c\/p\u003e \u003cp\u003e1.6 Summary and Further Research 33\u003c\/p\u003e \u003cp\u003eReferences 34\u003c\/p\u003e \u003cp\u003e\u003cb\u003e2 POWER-AWARE HIGH PERFORMANCE COMPUTING 39\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eRong Ge and Kirk W. Cameron\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e2.1 Introduction 39\u003c\/p\u003e \u003cp\u003e2.2 Background 41\u003c\/p\u003e \u003cp\u003e2.2.1 Current Hardware Technology and Power Consumption 41\u003c\/p\u003e \u003cp\u003e2.2.1.1 Processor power 41\u003c\/p\u003e \u003cp\u003e2.2.1.2 Memory subsystem power 42\u003c\/p\u003e \u003cp\u003e2.2.2 Performance 43\u003c\/p\u003e \u003cp\u003e2.2.3 Energy Efficiency 44\u003c\/p\u003e \u003cp\u003e2.3 Related Work 45\u003c\/p\u003e \u003cp\u003e2.3.1 Power Profiling 45\u003c\/p\u003e \u003cp\u003e2.3.1.1 Simulator-based power estimation 45\u003c\/p\u003e \u003cp\u003e2.3.1.2 Direct measurements 46\u003c\/p\u003e \u003cp\u003e2.3.1.3 Event-based estimation 46\u003c\/p\u003e \u003cp\u003e2.3.2 Performance Scalability on Power-Aware Systems 46\u003c\/p\u003e \u003cp\u003e2.3.3 Adaptive Power Allocation for Energy-Efficient Computing 47\u003c\/p\u003e \u003cp\u003e2.4 PowerPack: Fine-Grain Energy Profiling of HPC Applications 48\u003c\/p\u003e \u003cp\u003e2.4.1 Design and Implementation of PowerPack 48\u003c\/p\u003e \u003cp\u003e2.4.1.1 Overview 48\u003c\/p\u003e \u003cp\u003e2.4.1.2 Fine-grain systematic power measurement 50\u003c\/p\u003e \u003cp\u003e2.4.1.3 Automatic power profiling and code synchronization 51\u003c\/p\u003e \u003cp\u003e2.4.2 Power Profiles of HPC Applications and Systems 53\u003c\/p\u003e \u003cp\u003e2.4.2.1 Power distribution over components 53\u003c\/p\u003e \u003cp\u003e2.4.2.2 Power dynamics of applications 54\u003c\/p\u003e \u003cp\u003e2.4.2.3 Power bounds on HPC systems 55\u003c\/p\u003e \u003cp\u003e2.4.2.4 Power versus dynamic voltage and frequency scaling 57\u003c\/p\u003e \u003cp\u003e2.5 Power-Aware Speedup Model 59\u003c\/p\u003e \u003cp\u003e2.5.1 Power-Aware Speedup 59\u003c\/p\u003e \u003cp\u003e2.5.1.1 Sequential execution time for a single workload T1(w, f ) 60\u003c\/p\u003e \u003cp\u003e2.5.1.2 Sequential execution time for an ON-chip\/OFF-chip workload 60\u003c\/p\u003e \u003cp\u003e2.5.1.3 Parallel execution time on N processors for an ON-\/OFF-chip workload with DOP = i 61\u003c\/p\u003e \u003cp\u003e2.5.1.4 Power-aware speedup for DOP and ON-\/OFF-chip workloads 62\u003c\/p\u003e \u003cp\u003e2.5.2 Model Parametrization and Validation 63\u003c\/p\u003e \u003cp\u003e2.5.2.1 Coarse-grain parametrization and validation 64\u003c\/p\u003e \u003cp\u003e2.5.2.2 Fine-grain parametrization and validation 66\u003c\/p\u003e \u003cp\u003e2.6 Model Usages 69\u003c\/p\u003e \u003cp\u003e2.6.1 Identification of Optimal System Configurations 70\u003c\/p\u003e \u003cp\u003e2.6.2 PAS-Directed Energy-Driven Runtime Frequency Scaling 71\u003c\/p\u003e \u003cp\u003e2.7 Conclusion 73\u003c\/p\u003e \u003cp\u003eReferences 75\u003c\/p\u003e \u003cp\u003e\u003cb\u003e3 ENERGY EFFICIENCY IN HPC SYSTEMS 81\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eIvan Rodero and Manish Parashar\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e3.1 Introduction 81\u003c\/p\u003e \u003cp\u003e3.2 Background and Related Work 83\u003c\/p\u003e \u003cp\u003e3.2.1 CPU Power Management 83\u003c\/p\u003e \u003cp\u003e3.2.1.1 OS-level CPU power management 83\u003c\/p\u003e \u003cp\u003e3.2.1.2 Workload-level CPU power management 84\u003c\/p\u003e \u003cp\u003e3.2.1.3 Cluster-level CPU power management 84\u003c\/p\u003e \u003cp\u003e3.2.2 Component-Based Power Management 85\u003c\/p\u003e \u003cp\u003e3.2.2.1 Memory subsystem 85\u003c\/p\u003e \u003cp\u003e3.2.2.2 Storage subsystem 86\u003c\/p\u003e \u003cp\u003e3.2.3 Thermal-Conscious Power Management 87\u003c\/p\u003e \u003cp\u003e3.2.4 Power Management in Virtualized Datacenters 87\u003c\/p\u003e \u003cp\u003e3.3 Proactive, Component-Based Power Management 88\u003c\/p\u003e \u003cp\u003e3.3.1 Job Allocation Policies 88\u003c\/p\u003e \u003cp\u003e3.3.2 Workload Profiling 90\u003c\/p\u003e \u003cp\u003e3.4 Quantifying Energy Saving Possibilities 91\u003c\/p\u003e \u003cp\u003e3.4.1 Methodology 92\u003c\/p\u003e \u003cp\u003e3.4.2 Component-Level Power Requirements 92\u003c\/p\u003e \u003cp\u003e3.4.3 Energy Savings 94\u003c\/p\u003e \u003cp\u003e3.5 Evaluation of the Proposed Strategies 95\u003c\/p\u003e \u003cp\u003e3.5.1 Methodology 96\u003c\/p\u003e \u003cp\u003e3.5.2 Workloads 96\u003c\/p\u003e \u003cp\u003e3.5.3 Metrics 97\u003c\/p\u003e \u003cp\u003e3.6 Results 97\u003c\/p\u003e \u003cp\u003e3.7 Concluding Remarks 102\u003c\/p\u003e \u003cp\u003e3.8 Summary 103\u003c\/p\u003e \u003cp\u003eReferences 104\u003c\/p\u003e \u003cp\u003e\u003cb\u003e4 A STOCHASTIC FRAMEWORK FOR HIERARCHICAL SYSTEM-LEVEL POWER MANAGEMENT 109\u003cbr\u003e \u003c\/b\u003e\u003ci\u003ePeng Rong and Massoud Pedram\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e4.1 Introduction 109\u003c\/p\u003e \u003cp\u003e4.2 Related Work 111\u003c\/p\u003e \u003cp\u003e4.3 A Hierarchical DPM Architecture 113\u003c\/p\u003e \u003cp\u003e4.4 Modeling 114\u003c\/p\u003e \u003cp\u003e4.4.1 Model of the Application Pool 114\u003c\/p\u003e \u003cp\u003e4.4.2 Model of the Service Flow Control 118\u003c\/p\u003e \u003cp\u003e4.4.3 Model of the Simulated Service Provider 119\u003c\/p\u003e \u003cp\u003e4.4.4 Modeling Dependencies between SPs 120\u003c\/p\u003e \u003cp\u003e4.5 Policy Optimization 122\u003c\/p\u003e \u003cp\u003e4.5.1 Mathematical Formulation 122\u003c\/p\u003e \u003cp\u003e4.5.2 Optimal Time-Out Policy for Local Power Manager 123\u003c\/p\u003e \u003cp\u003e4.6 Experimental Results 125\u003c\/p\u003e \u003cp\u003e4.7 Conclusion 130\u003c\/p\u003e \u003cp\u003eReferences 130\u003c\/p\u003e \u003cp\u003e\u003cb\u003e5 ENERGY-EFFICIENT RESERVATION INFRASTRUCTURE FOR GRIDS, CLOUDS, AND NETWORKS 133\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eAnne-Ce´ cile Orgerie and Laurent Lefe` vre\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e5.1 Introduction 133\u003c\/p\u003e \u003cp\u003e5.2 Related Works 134\u003c\/p\u003e \u003cp\u003e5.2.1 Server and Data Center Power Management 135\u003c\/p\u003e \u003cp\u003e5.2.2 Node Optimizations 135\u003c\/p\u003e \u003cp\u003e5.2.3 Virtualization to Improve Energy Efficiency 136\u003c\/p\u003e \u003cp\u003e5.2.4 Energy Awareness in Wired Networking Equipment 136\u003c\/p\u003e \u003cp\u003e5.2.5 Synthesis 137\u003c\/p\u003e \u003cp\u003e5.3 ERIDIS: Energy-Efficient Reservation Infrastructure for Large-Scale Distributed Systems 138\u003c\/p\u003e \u003cp\u003e5.3.1 ERIDIS Architecture 138\u003c\/p\u003e \u003cp\u003e5.3.2 Management of the Resource Reservations 141\u003c\/p\u003e \u003cp\u003e5.3.3 Resource Management and On\/Off Algorithms 145\u003c\/p\u003e \u003cp\u003e5.3.4 Energy-Consumption Estimates 146\u003c\/p\u003e \u003cp\u003e5.3.5 Prediction Algorithms 146\u003c\/p\u003e \u003cp\u003e5.4 EARI: Energy-Aware Reservation Infrastructure for Data Centers and Grids 147\u003c\/p\u003e \u003cp\u003e5.4.1 EARI’s Architecture 147\u003c\/p\u003e \u003cp\u003e5.4.2 Validation of EARI on Experimental Grid Traces 147\u003c\/p\u003e \u003cp\u003e5.5 GOC: Green Open Cloud 149\u003c\/p\u003e \u003cp\u003e5.5.1 GOC’s Resource Manager Architecture 150\u003c\/p\u003e \u003cp\u003e5.5.2 Validation of the GOC Framework 152\u003c\/p\u003e \u003cp\u003e5.6 HERMES: High Level Energy-Aware Model for Bandwidth Reservation in End-To-End Networks 152\u003c\/p\u003e \u003cp\u003e5.6.1 HERMES’ Architecture 154\u003c\/p\u003e \u003cp\u003e5.6.2 The Reservation Process of HERMES 155\u003c\/p\u003e \u003cp\u003e5.6.3 Discussion 157\u003c\/p\u003e \u003cp\u003e5.7 Summary 158\u003c\/p\u003e \u003cp\u003eReferences 158\u003c\/p\u003e \u003cp\u003e\u003cb\u003e6 ENERGY-EFFICIENT JOB PLACEMENT ON CLUSTERS, GRIDS, AND CLOUDS 163\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eDamien Borgetto, Henri Casanova, Georges Da Costa, and Jean-Marc Pierson\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e6.1 Problem and Motivation 163\u003c\/p\u003e \u003cp\u003e6.1.1 Context 163\u003c\/p\u003e \u003cp\u003e6.1.2 Chapter Roadmap 164\u003c\/p\u003e \u003cp\u003e6.2 Energy-Aware Infrastructures 164\u003c\/p\u003e \u003cp\u003e6.2.1 Buildings 165\u003c\/p\u003e \u003cp\u003e6.2.2 Context-Aware Buildings 165\u003c\/p\u003e \u003cp\u003e6.2.3 Cooling 166\u003c\/p\u003e \u003cp\u003e6.3 Current Resource Management Practices 167\u003c\/p\u003e \u003cp\u003e6.3.1 Widely Used Resource Management Systems 167\u003c\/p\u003e \u003cp\u003e6.3.2 Job Requirement Description 169\u003c\/p\u003e \u003cp\u003e6.4 Scientific and Technical Challenges 170\u003c\/p\u003e \u003cp\u003e6.4.1 Theoretical Difficulties 170\u003c\/p\u003e \u003cp\u003e6.4.2 Technical Difficulties 170\u003c\/p\u003e \u003cp\u003e6.4.3 Controlling and Tuning Jobs 171\u003c\/p\u003e \u003cp\u003e6.5 Energy-Aware Job Placement Algorithms 172\u003c\/p\u003e \u003cp\u003e6.5.1 State of the Art 172\u003c\/p\u003e \u003cp\u003e6.5.2 Detailing One Approach 174\u003c\/p\u003e \u003cp\u003e6.6 Discussion 180\u003c\/p\u003e \u003cp\u003e6.6.1 Open Issues and Opportunities 180\u003c\/p\u003e \u003cp\u003e6.6.2 Obstacles for Adoption in Production 182\u003c\/p\u003e \u003cp\u003e6.7 Conclusion 183\u003c\/p\u003e \u003cp\u003eReferences 184\u003c\/p\u003e \u003cp\u003e\u003cb\u003e7 COMPARISON AND ANALYSIS OF GREEDY ENERGY-EFFICIENT SCHEDULING ALGORITHMS FOR COMPUTATIONAL GRIDS 189\u003cbr\u003e \u003c\/b\u003e\u003ci\u003ePeder Lindberg, James Leingang, Daniel Lysaker, Kashif Bilal, Samee Ullah Khan, Pascal Bouvry, Nasir Ghani, Nasro Min-Allah, and Juan Li\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e7.1 Introduction 189\u003c\/p\u003e \u003cp\u003e7.2 Problem Formulation 191\u003c\/p\u003e \u003cp\u003e7.2.1 The System Model 191\u003c\/p\u003e \u003cp\u003e7.2.1.1 PEs 191\u003c\/p\u003e \u003cp\u003e7.2.1.2 DVS 191\u003c\/p\u003e \u003cp\u003e7.2.1.3 Tasks 192\u003c\/p\u003e \u003cp\u003e7.2.1.4 Preliminaries 192\u003c\/p\u003e \u003cp\u003e7.2.2 Formulating the Energy-Makespan Minimization Problem 192\u003c\/p\u003e \u003cp\u003e7.3 Proposed Algorithms 193\u003c\/p\u003e \u003cp\u003e7.3.1 Greedy Heuristics 194\u003c\/p\u003e \u003cp\u003e7.3.1.1 Greedy heuristic scheduling algorithm 196\u003c\/p\u003e \u003cp\u003e7.3.1.2 Greedy-min 197\u003c\/p\u003e \u003cp\u003e7.3.1.3 Greedy-deadline 198\u003c\/p\u003e \u003cp\u003e7.3.1.4 Greedy-max 198\u003c\/p\u003e \u003cp\u003e7.3.1.5 MaxMin 199\u003c\/p\u003e \u003cp\u003e7.3.1.6 ObFun 199\u003c\/p\u003e \u003cp\u003e7.3.1.7 MinMin StdDev 202\u003c\/p\u003e \u003cp\u003e7.3.1.8 MinMax StdDev 202\u003c\/p\u003e \u003cp\u003e7.4 Simulations, Results, and Discussion 203\u003c\/p\u003e \u003cp\u003e7.4.1 Workload 203\u003c\/p\u003e \u003cp\u003e7.4.2 Comparative Results 204\u003c\/p\u003e \u003cp\u003e7.4.2.1 Small-size problems 204\u003c\/p\u003e \u003cp\u003e7.4.2.2 Large-size problems 206\u003c\/p\u003e \u003cp\u003e7.5 Related Works 211\u003c\/p\u003e \u003cp\u003e7.6 Conclusion 211\u003c\/p\u003e \u003cp\u003eReferences 212\u003c\/p\u003e \u003cp\u003e\u003cb\u003e8 TOWARD ENERGY-AWARE SCHEDULING USING MACHINE LEARNING 215\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eJosep LL. Berral, In˜ igo Goiri, Ramon Nou, Ferran Julia` , Josep O. Fito´ , Jordi Guitart, Ricard Gavalda´ , and Jordi Torres\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e8.1 Introduction 215\u003c\/p\u003e \u003cp\u003e8.1.1 Energetic Impact of the Cloud 216\u003c\/p\u003e \u003cp\u003e8.1.2 An Intelligent Way to Manage Data Centers 216\u003c\/p\u003e \u003cp\u003e8.1.3 Current Autonomic Computing Techniques 217\u003c\/p\u003e \u003cp\u003e8.1.4 Power-Aware Autonomic Computing 217\u003c\/p\u003e \u003cp\u003e8.1.5 State of the Art and Case Study 218\u003c\/p\u003e \u003cp\u003e8.2 Intelligent Self-Management 218\u003c\/p\u003e \u003cp\u003e8.2.1 Classical AI Approaches 219\u003c\/p\u003e \u003cp\u003e8.2.1.1 Heuristic algorithms 219\u003c\/p\u003e \u003cp\u003e8.2.1.2 AI planning 219\u003c\/p\u003e \u003cp\u003e8.2.1.3 Semantic techniques 219\u003c\/p\u003e \u003cp\u003e8.2.1.4 Expert systems and genetic algorithms 220\u003c\/p\u003e \u003cp\u003e8.2.2 Machine Learning Approaches 220\u003c\/p\u003e \u003cp\u003e8.2.2.1 Instance-based learning 221\u003c\/p\u003e \u003cp\u003e8.2.2.2 Reinforcement learning 222\u003c\/p\u003e \u003cp\u003e8.2.2.3 Feature and example selection 225\u003c\/p\u003e \u003cp\u003e8.3 Introducing Power-Aware Approaches 225\u003c\/p\u003e \u003cp\u003e8.3.1 Use of Virtualization 226\u003c\/p\u003e \u003cp\u003e8.3.2 Turning On and Off Machines 228\u003c\/p\u003e \u003cp\u003e8.3.3 Dynamic Voltage and Frequency Scaling 229\u003c\/p\u003e \u003cp\u003e8.3.4 Hybrid Nodes and Data Centers 230\u003c\/p\u003e \u003cp\u003e8.4 Experiences of Applying ML on Power-Aware Self-Management 230\u003c\/p\u003e \u003cp\u003e8.4.1 Case Study Approach 231\u003c\/p\u003e \u003cp\u003e8.4.2 Scheduling and Power Trade-Off 231\u003c\/p\u003e \u003cp\u003e8.4.3 Experimenting with Power-Aware Techniques 233\u003c\/p\u003e \u003cp\u003e8.4.4 Applying Machine Learning 236\u003c\/p\u003e \u003cp\u003e8.4.5 Conclusions from the Experiments 238\u003c\/p\u003e \u003cp\u003e8.5 Conclusions on Intelligent Power-Aware Self-Management 238\u003c\/p\u003e \u003cp\u003eReferences 240\u003c\/p\u003e \u003cp\u003e\u003cb\u003e9 ENERGY EFFICIENCY METRICS FOR DATA CENTERS 245\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eJavid Taheri and Albert Y. Zomaya\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e9.1 Introduction 245\u003c\/p\u003e \u003cp\u003e9.1.1 Background 245\u003c\/p\u003e \u003cp\u003e9.1.2 Data Center Energy Use 246\u003c\/p\u003e \u003cp\u003e9.1.3 Data Center Characteristics 246\u003c\/p\u003e \u003cp\u003e9.1.3.1 Electric power 247\u003c\/p\u003e \u003cp\u003e9.1.3.2 Heat removal 249\u003c\/p\u003e \u003cp\u003e9.1.4 Energy Efficiency 250\u003c\/p\u003e \u003cp\u003e9.2 Fundamentals of Metrics 250\u003c\/p\u003e \u003cp\u003e9.2.1 Demand and Constraints on Data Center Operators 250\u003c\/p\u003e \u003cp\u003e9.2.2 Metrics 251\u003c\/p\u003e \u003cp\u003e9.2.2.1 Criteria for good metrics 251\u003c\/p\u003e \u003cp\u003e9.2.2.2 Methodology 252\u003c\/p\u003e \u003cp\u003e9.2.2.3 Stability of metrics 252\u003c\/p\u003e \u003cp\u003e9.3 Data Center Energy Efficiency 252\u003c\/p\u003e \u003cp\u003e9.3.1 Holistic IT Efficiency Metrics 252\u003c\/p\u003e \u003cp\u003e9.3.1.1 Fixed versus proportional overheads 254\u003c\/p\u003e \u003cp\u003e9.3.1.2 Power versus energy 254\u003c\/p\u003e \u003cp\u003e9.3.1.3 Performance versus productivity 255\u003c\/p\u003e \u003cp\u003e9.3.2 Code of Conduct 256\u003c\/p\u003e \u003cp\u003e9.3.2.1 Environmental statement 256\u003c\/p\u003e \u003cp\u003e9.3.2.2 Problem statement 256\u003c\/p\u003e \u003cp\u003e9.3.2.3 Scope of the CoC 257\u003c\/p\u003e \u003cp\u003e9.3.2.4 Aims and objectives of CoC 258\u003c\/p\u003e \u003cp\u003e9.3.3 Power Use in Data Centers 259\u003c\/p\u003e \u003cp\u003e9.3.3.1 Data center IT power to utility power relationship 259\u003c\/p\u003e \u003cp\u003e9.3.3.2 Chiller efficiency and external temperature 260\u003c\/p\u003e \u003cp\u003e9.4 Available Metrics 260\u003c\/p\u003e \u003cp\u003e9.4.1 The Green Grid 261\u003c\/p\u003e \u003cp\u003e9.4.1.1 Power usage effectiveness (PUE) 261\u003c\/p\u003e \u003cp\u003e9.4.1.2 Data center efficiency (DCE) 262\u003c\/p\u003e \u003cp\u003e9.4.1.3 Data center infrastructure efficiency (DCiE) 262\u003c\/p\u003e \u003cp\u003e9.4.1.4 Data center productivity (DCP) 263\u003c\/p\u003e \u003cp\u003e9.4.2 McKinsey 263\u003c\/p\u003e \u003cp\u003e9.4.3 Uptime Institute 264\u003c\/p\u003e \u003cp\u003e9.4.3.1 Site infrastructure power overhead multiplier (SI-POM) 265\u003c\/p\u003e \u003cp\u003e9.4.3.2 IT hardware power overhead multiplier (H-POM) 266\u003c\/p\u003e \u003cp\u003e9.4.3.3 DC hardware compute load per unit of computing work done 266\u003c\/p\u003e \u003cp\u003e9.4.3.4 Deployed hardware utilization ratio (DH-UR) 266\u003c\/p\u003e \u003cp\u003e9.4.3.5 Deployed hardware utilization efficiency (DH-UE) 267\u003c\/p\u003e \u003cp\u003e9.5 Harmonizing Global Metrics for Data Center Energy Efficiency 267\u003c\/p\u003e \u003cp\u003eReferences 268\u003c\/p\u003e \u003cp\u003e\u003cb\u003e10 AUTONOMIC GREEN COMPUTING IN LARGE-SCALE DATA CENTERS 271\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eHaoting Luo, Bithika Khargharia, Salim Hariri, and Youssif Al-Nashif\u003cbr\u003e \u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e10.1 Introduction 271\u003c\/p\u003e \u003cp\u003e10.2 Related Technologies and Techniques 272\u003c\/p\u003e \u003cp\u003e10.2.1 Power Optimization Techniques in Data Centers 272\u003c\/p\u003e \u003cp\u003e10.2.2 Design Model 273\u003c\/p\u003e \u003cp\u003e10.2.3 Networks 274\u003c\/p\u003e \u003cp\u003e10.2.4 Data Center Power Distribution 275\u003c\/p\u003e \u003cp\u003e10.2.5 Data Center Power-Efficient Metrics 276\u003c\/p\u003e \u003cp\u003e10.2.6 Modeling Prototype and Testbed 277\u003c\/p\u003e \u003cp\u003e10.2.7 Green Computing 278\u003c\/p\u003e \u003cp\u003e10.2.8 Energy Proportional Computing 280\u003c\/p\u003e \u003cp\u003e10.2.9 Hardware Virtualization Technology 281\u003c\/p\u003e \u003cp\u003e10.2.10 Autonomic Computing 282\u003c\/p\u003e \u003cp\u003e10.3 Autonomic Green Computing: A Case Study 283\u003c\/p\u003e \u003cp\u003e10.3.1 Autonomic Management Platform 285\u003c\/p\u003e \u003cp\u003e10.3.1.1 Platform architecture 285\u003c\/p\u003e \u003cp\u003e10.3.1.2 DEVS-based modeling and simulation platform 285\u003c\/p\u003e \u003cp\u003e10.3.1.3 Workload generator 287\u003c\/p\u003e \u003cp\u003e10.3.2 Model Parameter Evaluation 288\u003c\/p\u003e \u003cp\u003e10.3.2.1 State transitioning overhead 288\u003c\/p\u003e \u003cp\u003e10.3.2.2 VM template evaluation 289\u003c\/p\u003e \u003cp\u003e10.3.2.3 Scalability analysis 291\u003c\/p\u003e \u003cp\u003e10.3.3 Autonomic Power Efficiency Management Algorithm (Performance Per Watt) 291\u003c\/p\u003e \u003cp\u003e10.3.4 Simulation Results and Evaluation 293\u003c\/p\u003e \u003cp\u003e10.3.4.1 Analysis of energy and performance trade-offs 296\u003c\/p\u003e \u003cp\u003e10.4 Conclusion and Future Directions 297\u003c\/p\u003e \u003cp\u003eReferences 298\u003c\/p\u003e \u003cp\u003e\u003cb\u003e11 ENERGY AND THERMAL AWARE SCHEDULING IN DATA CENTERS 301\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eGaurav Dhiman, Raid Ayoub, and Tajana S. Rosing\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e11.1 Introduction 301\u003c\/p\u003e \u003cp\u003e11.2 Related Work 302\u003c\/p\u003e \u003cp\u003e11.3 Intermachine Scheduling 305\u003c\/p\u003e \u003cp\u003e11.3.1 Performance and Power Profile of VMs 305\u003c\/p\u003e \u003cp\u003e11.3.2 Architecture 309\u003c\/p\u003e \u003cp\u003e11.3.2.1 vgnode 309\u003c\/p\u003e \u003cp\u003e11.3.2.2 vgxen 310\u003c\/p\u003e \u003cp\u003e11.3.2.3 vgdom 312\u003c\/p\u003e \u003cp\u003e11.3.2.4 vgserv 312\u003c\/p\u003e \u003cp\u003e11.4 Intramachine Scheduling 315\u003c\/p\u003e \u003cp\u003e11.4.1 Air-Forced Thermal Modeling and Cost 316\u003c\/p\u003e \u003cp\u003e11.4.2 Cooling Aware Dynamic Workload Scheduling 317\u003c\/p\u003e \u003cp\u003e11.4.3 Scheduling Mechanism 318\u003c\/p\u003e \u003cp\u003e11.4.4 Cooling Costs Predictor 319\u003c\/p\u003e \u003cp\u003e11.5 Evaluation 321\u003c\/p\u003e \u003cp\u003e11.5.1 Intermachine Scheduler (vGreen) 321\u003c\/p\u003e \u003cp\u003e11.5.2 Heterogeneous Workloads 323\u003c\/p\u003e \u003cp\u003e11.5.2.1 Comparison with DVFS policies 325\u003c\/p\u003e \u003cp\u003e11.5.2.2 Homogeneous workloads 328\u003c\/p\u003e \u003cp\u003e11.5.3 Intramachine Scheduler (Cool and Save) 328\u003c\/p\u003e \u003cp\u003e11.5.3.1 Results 331\u003c\/p\u003e \u003cp\u003e11.5.3.2 Overhead of CAS 333\u003c\/p\u003e \u003cp\u003e11.6 Conclusion 333\u003c\/p\u003e \u003cp\u003eReferences 334\u003c\/p\u003e \u003cp\u003e\u003cb\u003e12 QOS-AWARE POWER MANAGEMENT IN DATA CENTERS 339\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eJiayu Gong and Cheng-Zhong Xu\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e12.1 Introduction 339\u003c\/p\u003e \u003cp\u003e12.2 Problem Classification 340\u003c\/p\u003e \u003cp\u003e12.2.1 Objective and Constraint 340\u003c\/p\u003e \u003cp\u003e12.2.2 Scope and Time Granularities 340\u003c\/p\u003e \u003cp\u003e12.2.3 Methodology 341\u003c\/p\u003e \u003cp\u003e12.2.4 Power Management Mechanism 342\u003c\/p\u003e \u003cp\u003e12.3 Energy Efficiency 344\u003c\/p\u003e \u003cp\u003e12.3.1 Energy-Efficiency Metrics 344\u003c\/p\u003e \u003cp\u003e12.3.2 Improving Energy Efficiency 346\u003c\/p\u003e \u003cp\u003e12.3.2.1 Energy minimization with performance guarantee 346\u003c\/p\u003e \u003cp\u003e12.3.2.2 Performance maximization under power budget 348\u003c\/p\u003e \u003cp\u003e12.3.2.3 Trade-off between power and performance 348\u003c\/p\u003e \u003cp\u003e12.3.3 Energy-Proportional Computing 350\u003c\/p\u003e \u003cp\u003e12.4 Power Capping 351\u003c\/p\u003e \u003cp\u003e12.5 Conclusion 353\u003c\/p\u003e \u003cp\u003eReferences 356\u003c\/p\u003e \u003cp\u003e\u003cb\u003e13 ENERGY-EFFICIENT STORAGE SYSTEMS FOR DATA CENTERS 361\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eSudhanva Gurumurthi and Anand Sivasubramaniam\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e13.1 Introduction 361\u003c\/p\u003e \u003cp\u003e13.2 Disk Drive Operation and Disk Power 362\u003c\/p\u003e \u003cp\u003e13.2.1 An Overview of Disk Drives 362\u003c\/p\u003e \u003cp\u003e13.2.2 Sources of Disk Power Consumption 363\u003c\/p\u003e \u003cp\u003e13.2.3 Disk Activity and Power Consumption 365\u003c\/p\u003e \u003cp\u003e13.3 Disk and Storage Power Reduction Techniques 366\u003c\/p\u003e \u003cp\u003e13.3.1 Exploiting the STANDBY State 368\u003c\/p\u003e \u003cp\u003e13.3.2 Reducing Seek Activity 369\u003c\/p\u003e \u003cp\u003e13.3.3 Achieving Energy Proportionality 369\u003c\/p\u003e \u003cp\u003e13.3.3.1 Hardware approaches 369\u003c\/p\u003e \u003cp\u003e13.3.3.2 Software approaches 370\u003c\/p\u003e \u003cp\u003e13.4 Using Nonvolatile Memory and Solid-State Disks 371\u003c\/p\u003e \u003cp\u003e13.5 Conclusions 372\u003c\/p\u003e \u003cp\u003eReferences 373\u003c\/p\u003e \u003cp\u003e\u003cb\u003e14 AUTONOMIC ENERGY\/PERFORMANCE OPTIMIZATIONS FOR MEMORY IN SERVERS 377\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eBithika Khargharia and Mazin Yousif\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e14.1 Introduction 378\u003c\/p\u003e \u003cp\u003e14.2 Classifications of Dynamic Power Management Techniques 380\u003c\/p\u003e \u003cp\u003e14.2.1 Heuristic and Predictive Techniques 380\u003c\/p\u003e \u003cp\u003e14.2.2 QoS and Energy Trade-Offs 381\u003c\/p\u003e \u003cp\u003e14.3 Applications of Dynamic Power Management (DPM) 382\u003c\/p\u003e \u003cp\u003e14.3.1 Power Management of System Components in Isolation 382\u003c\/p\u003e \u003cp\u003e14.3.2 Joint Power Management of System Components 383\u003c\/p\u003e \u003cp\u003e14.3.3 Holistic System-Level Power Management 383\u003c\/p\u003e \u003cp\u003e14.4 Autonomic Power and Performance Optimization of Memory Subsystems in Server Platforms 384\u003c\/p\u003e \u003cp\u003e14.4.1 Adaptive Memory Interleaving Technique for Power and Performance Management 384\u003c\/p\u003e \u003cp\u003e14.4.1.1 Formulating the optimization problem 386\u003c\/p\u003e \u003cp\u003e14.4.1.2 Memory appflow 389\u003c\/p\u003e \u003cp\u003e14.4.2 Industry Techniques 389\u003c\/p\u003e \u003cp\u003e14.4.2.1 Enhancements in memory hardware design 390\u003c\/p\u003e \u003cp\u003e14.4.2.2 Adding more operating states 390\u003c\/p\u003e \u003cp\u003e14.4.2.3 Faster transition to and from low power states 390\u003c\/p\u003e \u003cp\u003e14.4.2.4 Memory consolidation 390\u003c\/p\u003e \u003cp\u003e14.5 Conclusion 391\u003c\/p\u003e \u003cp\u003eReferences 391\u003c\/p\u003e \u003cp\u003e\u003cb\u003e15 ROD: A PRACTICAL APPROACH TO IMPROVING RELIABILITY OF ENERGY-EFFICIENT PARALLEL \u003ci\u003eDISK SYSTEMS 395\u003cbr\u003e \u003c\/i\u003e\u003c\/b\u003e\u003ci\u003eShu Yin, Xiaojun Ruan, Adam Manzanares, and Xiao Qin\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e15.1 Introduction 395\u003c\/p\u003e \u003cp\u003e15.2 Modeling Reliability of Energy-Efficient Parallel Disks 396\u003c\/p\u003e \u003cp\u003e15.2.1 The MINT Model 396\u003c\/p\u003e \u003cp\u003e15.2.1.1 Disk utilization 398\u003c\/p\u003e \u003cp\u003e15.2.1.2 Temperature 398\u003c\/p\u003e \u003cp\u003e15.2.1.3 Power-state transition frequency 399\u003c\/p\u003e \u003cp\u003e15.2.1.4 Single disk reliability model 399\u003c\/p\u003e \u003cp\u003e15.2.2 MAID, Massive Arrays of Idle Disks 400\u003c\/p\u003e \u003cp\u003e15.3 Improving Reliability of MAID via Disk Swapping 401\u003c\/p\u003e \u003cp\u003e15.3.1 Improving Reliability of Cache Disks in MAID 401\u003c\/p\u003e \u003cp\u003e15.3.2 Swapping Disks Multiple Times 404\u003c\/p\u003e \u003cp\u003e15.4 Experimental Results and Evaluation 405\u003c\/p\u003e \u003cp\u003e15.4.1 Experimental Setup 405\u003c\/p\u003e \u003cp\u003e15.4.2 Disk Utilization 406\u003c\/p\u003e \u003cp\u003e15.4.3 The Single Disk Swapping Strategy 406\u003c\/p\u003e \u003cp\u003e15.4.4 The Multiple Disk Swapping Strategy 409\u003c\/p\u003e \u003cp\u003e15.5 Related Work 411\u003c\/p\u003e \u003cp\u003e15.6 Conclusions 412\u003c\/p\u003e \u003cp\u003eReferences 413\u003c\/p\u003e \u003cp\u003e\u003cb\u003e16 EMBRACING THE MEMORY AND I\/O WALLS FOR ENERGY-EFFICIENT SCIENTIFIC COMPUTING 417\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eChung-Hsing Hsu and Wu-Chun Feng\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e16.1 Introduction 417\u003c\/p\u003e \u003cp\u003e16.2 Background and Related Work 420\u003c\/p\u003e \u003cp\u003e16.2.1 DVFS-Enabled Processors 420\u003c\/p\u003e \u003cp\u003e16.2.2 DVFS Scheduling Algorithms 421\u003c\/p\u003e \u003cp\u003e16.2.3 Memory-Aware, Interval-Based Algorithms 422\u003c\/p\u003e \u003cp\u003e16.3 β-Adaptation: A New DVFS Algorithm 423\u003c\/p\u003e \u003cp\u003e16.3.1 The Compute-Boundedness Metric, β 423\u003c\/p\u003e \u003cp\u003e16.3.2 The Frequency Calculating Formula, f ∗ 424\u003c\/p\u003e \u003cp\u003e16.3.3 The Online β Estimation 425\u003c\/p\u003e \u003cp\u003e16.3.4 Putting It All Together 427\u003c\/p\u003e \u003cp\u003e16.4 Algorithm Effectiveness 429\u003c\/p\u003e \u003cp\u003e16.4.1 A Comparison to Other DVFS Algorithms 429\u003c\/p\u003e \u003cp\u003e16.4.2 Frequency Emulation 432\u003c\/p\u003e \u003cp\u003e16.4.3 The Minimum Dependence to the PMU 436\u003c\/p\u003e \u003cp\u003e16.5 Conclusions and Future Work 438\u003c\/p\u003e \u003cp\u003eReferences 439\u003c\/p\u003e \u003cp\u003e\u003cb\u003e17 MULTIPLE FREQUENCY SELECTION IN DVFS-ENABLED PROCESSORS TO MINIMIZE ENERGY CONSUMPTION 443\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eNikzad Babaii Rizvandi, Albert Y. Zomaya, Young Choon Lee, Ali Javadzadeh Boloori, and Javid Taheri\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e17.1 Introduction 443\u003c\/p\u003e \u003cp\u003e17.2 Energy Efficiency in HPC Systems 444\u003c\/p\u003e \u003cp\u003e17.3 Exploitation of Dynamic Voltage–Frequency Scaling 446\u003c\/p\u003e \u003cp\u003e17.3.1 Independent Slack Reclamation 446\u003c\/p\u003e \u003cp\u003e17.3.2 Integrated Schedule Generation 447\u003c\/p\u003e \u003cp\u003e17.4 Preliminaries 448\u003c\/p\u003e \u003cp\u003e17.4.1 System and Application Models 448\u003c\/p\u003e \u003cp\u003e17.4.2 Energy Model 448\u003c\/p\u003e \u003cp\u003e17.5 Energy-Aware Scheduling via DVFS 450\u003c\/p\u003e \u003cp\u003e17.5.1 Optimum Continuous Frequency 450\u003c\/p\u003e \u003cp\u003e17.5.2 Reference Dynamic Voltage–Frequency Scaling (RDVFS) 451\u003c\/p\u003e \u003cp\u003e17.5.3 Maximum-Minimum-Frequency for Dynamic Voltage–Frequency Scaling (MMF-DVFS) 452\u003c\/p\u003e \u003cp\u003e17.5.4 Multiple Frequency Selection for Dynamic Voltage–Frequency Scaling (MFS-DVFS) 453\u003c\/p\u003e \u003cp\u003e17.5.4.1 Task eligibility 454\u003c\/p\u003e \u003cp\u003e17.6 Experimental Results 456\u003c\/p\u003e \u003cp\u003e17.6.1 Simulation Settings 456\u003c\/p\u003e \u003cp\u003e17.6.2 Results 458\u003c\/p\u003e \u003cp\u003e17.7 Conclusion 461\u003c\/p\u003e \u003cp\u003eReferences 461\u003c\/p\u003e \u003cp\u003e\u003cb\u003e18 THE PARAMOUNTCY OF RECONFIGURABLE COMPUTING 465\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eReiner Hartenstein\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e18.1 Introduction 465\u003c\/p\u003e \u003cp\u003e18.2 Why Computers are Important 466\u003c\/p\u003e \u003cp\u003e18.2.1 Computing for a Sustainable Environment 470\u003c\/p\u003e \u003cp\u003e18.3 Performance Progress Stalled 472\u003c\/p\u003e \u003cp\u003e18.3.1 Unaffordable Energy Consumption of Computing 473\u003c\/p\u003e \u003cp\u003e18.3.2 Crashing into the Programming Wall 475\u003c\/p\u003e \u003cp\u003e18.4 The Tail is Wagging the Dog (Accelerators) 488\u003c\/p\u003e \u003cp\u003e18.4.1 Hardwired Accelerators 489\u003c\/p\u003e \u003cp\u003e18.4.2 Programmable Accelerators 490\u003c\/p\u003e \u003cp\u003e18.5 Reconfigurable Computing 494\u003c\/p\u003e \u003cp\u003e18.5.1 Speedup Factors by FPGAs 498\u003c\/p\u003e \u003cp\u003e18.5.2 The Reconfigurable Computing Paradox 501\u003c\/p\u003e \u003cp\u003e18.5.3 Saving Energy by Reconfigurable Computing 505\u003c\/p\u003e \u003cp\u003e18.5.3.1 Traditional green computing 506\u003c\/p\u003e \u003cp\u003e18.5.3.2 The role of graphics processors 507\u003c\/p\u003e \u003cp\u003e18.5.3.3 Wintel versus ARM 508\u003c\/p\u003e \u003cp\u003e18.5.4 Reconfigurable Computing is the Silver Bullet 511\u003c\/p\u003e \u003cp\u003e18.5.4.1 A new world model of computing 511\u003c\/p\u003e \u003cp\u003e18.5.5 The Twin-Paradigm Approach to Tear Down the Wall 514\u003c\/p\u003e \u003cp\u003e18.5.6 A Mass Movement Needed as Soon as Possible 517\u003c\/p\u003e \u003cp\u003e18.5.6.1 Legacy software from the mainframe age 518\u003c\/p\u003e \u003cp\u003e18.5.7 How to Reinvent Computing 519 18.6 Conclusions 526\u003c\/p\u003e \u003cp\u003eReferences 529\u003c\/p\u003e \u003cp\u003e\u003cb\u003e19 WORKLOAD CLUSTERING FOR INCREASING ENERGY SAVINGS ON EMBEDDED MPSOCS 549\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eOzcan Ozturk, Mahmut Kandemir, and Sri Hari Krishna Narayanan\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e19.1 Introduction 549\u003c\/p\u003e \u003cp\u003e19.2 Embedded MPSoC Architecture, Execution Model, and Related Work 550\u003c\/p\u003e \u003cp\u003e19.3 Our Approach 551\u003c\/p\u003e \u003cp\u003e19.3.1 Overview 551\u003c\/p\u003e \u003cp\u003e19.3.2 Technical Details and Problem Formulation 553\u003c\/p\u003e \u003cp\u003e19.3.2.1 System and job model 553\u003c\/p\u003e \u003cp\u003e19.3.2.2 Mathematical programing model 554\u003c\/p\u003e \u003cp\u003e19.3.2.3 Example 557\u003c\/p\u003e \u003cp\u003e19.4 Experimental Evaluation 560\u003c\/p\u003e \u003cp\u003e19.5 Conclusions 564\u003c\/p\u003e \u003cp\u003eReferences 565\u003c\/p\u003e \u003cp\u003e\u003cb\u003e20 ENERGY-EFFICIENT INTERNET INFRASTRUCTURE 567\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eWeirong Jiang and Viktor K. Prasanna\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e20.1 Introduction 567\u003c\/p\u003e \u003cp\u003e20.1.1 Performance Challenges 568\u003c\/p\u003e \u003cp\u003e20.1.2 Existing Packet Forwarding Approaches 570\u003c\/p\u003e \u003cp\u003e20.1.2.1 Software approaches 570\u003c\/p\u003e \u003cp\u003e20.1.2.2 Hardware approaches 571\u003c\/p\u003e \u003cp\u003e20.2 SRAM-Based Pipelined IP Lookup Architectures: Alternative to TCAMs 571\u003c\/p\u003e \u003cp\u003e20.3 Data Structure Optimization for Power Efficiency 573\u003c\/p\u003e \u003cp\u003e20.3.1 Problem Formulation 574\u003c\/p\u003e \u003cp\u003e20.3.1.1 Non-pipelined and pipelined engines 574\u003c\/p\u003e \u003cp\u003e20.3.1.2 Power function of SRAM 575\u003c\/p\u003e \u003cp\u003e20.3.2 Special Case: Uniform Stride 576\u003c\/p\u003e \u003cp\u003e20.3.3 Dynamic Programming 576\u003c\/p\u003e \u003cp\u003e20.3.4 Performance Evaluation 577\u003c\/p\u003e \u003cp\u003e20.3.4.1 Results for non-pipelined architecture 578\u003c\/p\u003e \u003cp\u003e20.3.4.2 Results for pipelined architecture 578\u003c\/p\u003e \u003cp\u003e20.4 Architectural Optimization to Reduce Dynamic Power Dissipation 580\u003c\/p\u003e \u003cp\u003e20.4.1 Analysis and Motivation 581\u003c\/p\u003e \u003cp\u003e20.4.1.1 Traffic locality 582\u003c\/p\u003e \u003cp\u003e20.4.1.2 Traffic rate variation 582\u003c\/p\u003e \u003cp\u003e20.4.1.3 Access frequency on different stages 583\u003c\/p\u003e \u003cp\u003e20.4.2 Architecture-Specific Techniques 583\u003c\/p\u003e \u003cp\u003e20.4.2.1 Inherent caching 584\u003c\/p\u003e \u003cp\u003e20.4.2.2 Local clocking 584\u003c\/p\u003e \u003cp\u003e20.4.2.3 Fine-grained memory enabling 585\u003c\/p\u003e \u003cp\u003e20.4.3 Performance Evaluation 585\u003c\/p\u003e \u003cp\u003e20.5 Related Work 588\u003c\/p\u003e \u003cp\u003e20.6 Summary 589\u003c\/p\u003e \u003cp\u003eReferences 589\u003c\/p\u003e \u003cp\u003e\u003cb\u003e21 DEMAND RESPONSE IN THE SMART GRID: A DISTRIBUTED COMPUTING PERSPECTIVE 593\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eChen Wang and Martin De Groot\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e21.1 Introduction 593\u003c\/p\u003e \u003cp\u003e21.2 Demand Response 595\u003c\/p\u003e \u003cp\u003e21.2.1 Existing Demand Response Programs 595\u003c\/p\u003e \u003cp\u003e21.2.2 Demand Response Supported by the Smart Grid 597\u003c\/p\u003e \u003cp\u003e21.3 Demand Response as a Distributed System 600\u003c\/p\u003e \u003cp\u003e21.3.1 An Overlay Network for Demand Response 600\u003c\/p\u003e \u003cp\u003e21.3.2 Event Driven Demand Response 602\u003c\/p\u003e \u003cp\u003e21.3.3 Cost Driven Demand Response 604\u003c\/p\u003e \u003cp\u003e21.3.4 A Decentralized Demand Response Framework 609\u003c\/p\u003e \u003cp\u003e21.3.5 Accountability of Coordination Decision Making 610\u003c\/p\u003e \u003cp\u003e21.4 Summary 611\u003c\/p\u003e \u003cp\u003eReferences 611\u003c\/p\u003e \u003cp\u003e\u003cb\u003e22 RESOURCE MANAGEMENT FOR DISTRIBUTED MOBILE COMPUTING 615\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eJong-Kook Kim\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e22.1 Introduction 615\u003c\/p\u003e \u003cp\u003e22.2 Single-Hop Energy-Constrained Environment 617\u003c\/p\u003e \u003cp\u003e22.2.1 System Model 617\u003c\/p\u003e \u003cp\u003e22.2.2 Related Work 620\u003c\/p\u003e \u003cp\u003e22.2.3 Heuristic Descriptions 621\u003c\/p\u003e \u003cp\u003e22.2.3.1 Mapping event 621\u003c\/p\u003e \u003cp\u003e22.2.3.2 Scheduling communications 621\u003c\/p\u003e \u003cp\u003e22.2.3.3 Opportunistic load balancing and minimum energy greedy heuristics 622\u003c\/p\u003e \u003cp\u003e22.2.3.4 ME-MC heuristic 622\u003c\/p\u003e \u003cp\u003e22.2.3.5 ME-ME heuristic 624\u003c\/p\u003e \u003cp\u003e22.2.3.6 CRME heuristic 625\u003c\/p\u003e \u003cp\u003e22.2.3.7 Originator and random 626\u003c\/p\u003e \u003cp\u003e22.2.3.8 Upper bound 626\u003c\/p\u003e \u003cp\u003e22.2.4 Simulation Model 628\u003c\/p\u003e \u003cp\u003e22.2.5 Results 630\u003c\/p\u003e \u003cp\u003e22.2.6 Summary 634\u003c\/p\u003e \u003cp\u003e22.3 Multihop Distributed Mobile Computing Environment 635\u003c\/p\u003e \u003cp\u003e22.3.1 The Multihop System Model 635\u003c\/p\u003e \u003cp\u003e22.3.2 Energy-Aware Routing Protocol 636\u003c\/p\u003e \u003cp\u003e22.3.2.1 Overview 636\u003c\/p\u003e \u003cp\u003e22.3.2.2 DSDV 637\u003c\/p\u003e \u003cp\u003e22.3.2.3 DSDV remaining energy 637\u003c\/p\u003e \u003cp\u003e22.3.2.4 DSDV-energy consumption per remaining energy 637\u003c\/p\u003e \u003cp\u003e22.3.3 Heuristic Description 638\u003c\/p\u003e \u003cp\u003e22.3.3.1 Random 638\u003c\/p\u003e \u003cp\u003e22.3.3.2 Estimated minimum total energy (EMTE) 638\u003c\/p\u003e \u003cp\u003e22.3.3.3 K-percent-speed (KPS) and K-percent-energy (KPE) 639\u003c\/p\u003e \u003cp\u003e22.3.3.4 Energy ratio and distance (ERD) 639\u003c\/p\u003e \u003cp\u003e22.3.3.5 ETC and distance (ETCD) 640\u003c\/p\u003e \u003cp\u003e22.3.3.6 Minimum execution time (MET) 640\u003c\/p\u003e \u003cp\u003e22.3.3.7 Minimum completion time (MCT) and minimum completion time with DVS (MCT-DVS) 640\u003c\/p\u003e \u003cp\u003e22.3.3.8 Switching algorithm (SA) 640\u003c\/p\u003e \u003cp\u003e22.3.4 Simulation Model 641\u003c\/p\u003e \u003cp\u003e22.3.5 Results 643\u003c\/p\u003e \u003cp\u003e22.3.5.1 Distributed resource management 643\u003c\/p\u003e \u003cp\u003e22.3.5.2 Energy-aware protocol 644\u003c\/p\u003e \u003cp\u003e22.3.6 Summary 644\u003c\/p\u003e \u003cp\u003e22.4 Future Work 647\u003c\/p\u003e \u003cp\u003eReferences 647\u003c\/p\u003e \u003cp\u003e\u003cb\u003e23 AN ENERGY-AWARE FRAMEWORK FOR MOBILE DATA MINING 653\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eCarmela Comito, Domenico Talia, and Paolo Trunfio\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e23.1 Introduction 653\u003c\/p\u003e \u003cp\u003e23.2 System Architecture 654\u003c\/p\u003e \u003cp\u003e23.3 Mobile Device Components 657\u003c\/p\u003e \u003cp\u003e23.4 Energy Model 659\u003c\/p\u003e \u003cp\u003e23.5 Clustering Scheme 664\u003c\/p\u003e \u003cp\u003e23.5.1 Clustering the M2M Architecture 666\u003c\/p\u003e \u003cp\u003e23.6 Conclusion 670\u003c\/p\u003e \u003cp\u003eReferences 670\u003c\/p\u003e \u003cp\u003e\u003cb\u003e24 ENERGY AWARENESS AND EFFICIENCY IN WIRELESS SENSOR NETWORKS: FROM PHYSICAL DEVICES TO THE COMMUNICATION LINK 673\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eFla´ via C. Delicato and Paulo F. Pires\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e24.1 Introduction 673\u003c\/p\u003e \u003cp\u003e24.2 WSN and Power Dissipation Models 676\u003c\/p\u003e \u003cp\u003e24.2.1 Network and Node Architecture 676\u003c\/p\u003e \u003cp\u003e24.2.2 Sources of Power Dissipation in WSNs 679\u003c\/p\u003e \u003cp\u003e24.3 Strategies for Energy Optimization 683\u003c\/p\u003e \u003cp\u003e24.3.1 Intranode Level 684\u003c\/p\u003e \u003cp\u003e24.3.1.1 Duty cycling 685\u003c\/p\u003e \u003cp\u003e24.3.1.2 Adaptive sensing 691\u003c\/p\u003e \u003cp\u003e24.3.1.3 Dynamic voltage scale (DVS) 693\u003c\/p\u003e \u003cp\u003e24.3.1.4 OS task scheduling 694\u003c\/p\u003e \u003cp\u003e24.3.2 Internode Level 695\u003c\/p\u003e \u003cp\u003e24.3.2.1 Transmission power control 695\u003c\/p\u003e \u003cp\u003e24.3.2.2 Dynamic modulation scaling 696\u003c\/p\u003e \u003cp\u003e24.3.2.3 Link layer optimizations 698\u003c\/p\u003e \u003cp\u003e24.4 Final Remarks 701\u003c\/p\u003e \u003cp\u003eReferences 702\u003c\/p\u003e \u003cp\u003e\u003cb\u003e25 NETWORK-WIDE STRATEGIES FOR ENERGY EFFICIENCY IN WIRELESS SENSOR NETWORKS 709\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eFla´ via C. Delicato and Paulo F. Pires\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e25.1 Introduction 709\u003c\/p\u003e \u003cp\u003e25.2 Data Link Layer 711\u003c\/p\u003e \u003cp\u003e25.2.1 Topology Control Protocols 712\u003c\/p\u003e \u003cp\u003e25.2.2 Energy-Efficient MAC Protocols 714\u003c\/p\u003e \u003cp\u003e25.2.2.1 Scheduled MAC protocols in WSNs 716\u003c\/p\u003e \u003cp\u003e25.2.2.2 Contention-based MAC protocols 717\u003c\/p\u003e \u003cp\u003e25.3 Network Layer 719\u003c\/p\u003e \u003cp\u003e25.3.1 Flat and Hierarchical Protocols 722\u003c\/p\u003e \u003cp\u003e25.4 Transport Layer 725\u003c\/p\u003e \u003cp\u003e25.5 Application Layer 729\u003c\/p\u003e \u003cp\u003e25.5.1 Task Scheduling 729\u003c\/p\u003e \u003cp\u003e25.5.2 Data Aggregation and Data Fusion in WSNs 733\u003c\/p\u003e \u003cp\u003e25.5.2.1 Approaches of data fusion for energy efficiency 735\u003c\/p\u003e \u003cp\u003e25.5.2.2 Data aggregation strategies 736\u003c\/p\u003e \u003cp\u003e25.6 Final Remarks 740\u003c\/p\u003e \u003cp\u003eReferences 741\u003c\/p\u003e \u003cp\u003e\u003cb\u003e26 ENERGY MANAGEMENT IN HETEROGENEOUS WIRELESS HEALTH CARE NETWORKS 751\u003cbr\u003e \u003c\/b\u003e\u003ci\u003eNima Nikzad, Priti Aghera, Piero Zappi, and Tajana S. Rosing\u003c\/i\u003e\u003c\/p\u003e \u003cp\u003e26.1 Introduction 751\u003c\/p\u003e \u003cp\u003e26.2 System Model 753\u003c\/p\u003e \u003cp\u003e26.2.1 Health Monitoring Task Model 753\u003c\/p\u003e \u003cp\u003e26.3 Collaborative Distributed Environmental Sensing 755\u003c\/p\u003e \u003cp\u003e26.3.1 Node Neighborhood and Localization Rate 757\u003c\/p\u003e \u003cp\u003e26.3.2 Energy Ratio and Sensing Rate 758\u003c\/p\u003e \u003cp\u003e26.3.3 Duty Cycling and Prediction 759\u003c\/p\u003e \u003cp\u003e26.4 Task Assignment in a Body Area Network 760\u003c\/p\u003e \u003cp\u003e26.4.1 Optimal Task Assignment 760\u003c\/p\u003e \u003cp\u003e26.4.2 Dynamic Task Assignment 762\u003c\/p\u003e \u003cp\u003e26.4.2.1 DynAGreen algorithm 763\u003c\/p\u003e \u003cp\u003e26.4.2.2 DynAGreenLife algorithm 768\u003c\/p\u003e \u003cp\u003e26.5 Results 771\u003c\/p\u003e \u003cp\u003e26.5.1 Collaborative Sensing 771\u003c\/p\u003e \u003cp\u003e26.5.1.1 Results 772\u003c\/p\u003e \u003cp\u003e26.5.2 Dynamic Task Assignment 776\u003c\/p\u003e \u003cp\u003e26.5.2.1 Performance in static conditions 777\u003c\/p\u003e \u003cp\u003e26.5.2.2 Dynamic adaptability 780\u003c\/p\u003e \u003cp\u003e26.6 Conclusion 784\u003c\/p\u003e \u003cp\u003eReferences 785\u003c\/p\u003e \u003cp\u003eINDEX 787\u003c\/p\u003e\u003c\/font\u003e\u003c\/p\u003e\r\n\r\n\u003cp\u003e\u003cfont size=\"3\"\u003eSubject Areas: Computer networking \u0026amp; communications [\u003ca title=\"See our other books on Computer networking \u0026amp; communications\" href=\"https:\/\/freshlyprintedbooks.co.uk\/search?q=%22Computer%20networking%20\u0026amp;%20communications%20%5BUT%5D%22\"\u003eUT\u003c\/a\u003e]\u003c\/font\u003e\u003c\/p\u003e\r\n\r\n\r\n\u003c\/font\u003e","brand":"Wiley-IEEE Computer Society Pr","offers":[{"title":"Brand New","offer_id":52278117564696,"sku":"9780470908754","price":96.99,"currency_code":"GBP","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0730\/2037\/5320\/files\/9780470908754.jpg?v=1781458410","url":"https:\/\/freshlyprintedbooks.co.uk\/products\/energy-efficient-distributed-computing-systems-hardback-9780470908754","provider":"Freshly Printed Books","version":"1.0","type":"link"}