Energy consumption is one of the main limiting factors for the design of ultrascale infrastructures. Multi-level hardware and software optimizations must be designed and explored in order to reduce energy consumption for these largescale equipment. This chapter addresses the issue of energy efficiency of ultrascale systems in front of other quality metrics. The goal of this chapter is to explore the design of metrics, analysis, frameworks and tools for putting energy awareness and energy efficiency at the next stage. Significant emphasis will be placed on the idea of “energy complexity,” reflecting the synergies between energy efficiency and quality of service, resilience and performance, by studying computation power, communication/data sharing power, data access power, algorithm energy consumption, etc.

Chapter Contents:

• 5.1 Energy modeling and simulation
• 5.1.1 Related work
• 5.1.2 Actor definition
• 5.1.3 Energy modeling
• 5.1.4 Power modeling
• 5.1.4.1 CPU modeling
• 5.1.4.2 Network modeling
• 5.1.4.3 I/O modeling
• 5.1.4.4 Thermal modeling
• 5.1.5 Runtime modeling
• 5.1.5.1 Workload modeling
• 5.1.5.2 Model evaluation
• 5.1.6 Simulation
• 5.1.7 Issues and open challenges
• 5.1.8 Conclusion
• Acknowledgments
• 5.2 Evaluation of renewable energy usage for an Exascale computer
• 5.2.1 Renewable energy modeling
• 5.2.1.1 Solar power model
• 5.2.1.2 Wind power model
• 5.2.1.3 Renewable quantity model
• 5.2.1.4 Minimum percentage supply
• 5.2.1.5 Cost evaluation
• 5.2.2 Battery modeling
• 5.2.2.1 Simulator tool
• 5.2.3 Geographical location competitiveness
• 5.2.3.1 Solar energy availability
• 5.2.3.2 Battery size for three different locations
• 5.2.4 Renewable energy and Exascale computing
• 5.2.5 Policy for renewable-powered data centers
• 5.2.5.1 European Commission
• 5.2.5.2 USA
• 5.2.5.3 Collaborative projects
• 5.2.5.4 Companies
• 5.3 An introduction to cooling in data centers
• 5.3.1 Overview of data center cooling methodologies
• 5.3.2 Cost estimation
• 5.3.3 Simulation tools
• 5.3.3.1 Electrical load simulation and prediction
• 5.3.3.2 Thermofluid simulations
• 5.3.4 Future challenges
• 5.3.5 Conclusion
• Acknowledgments
• 5.4 A full-cost model for estimating the energy consumption of computing infrastructures
• 5.4.1 Total cost of ownership analysis for a medium-size academic HPC facility
• Amazon EC2 instances for HPC and equivalent cost models
• 5.4.2 Energy cost models for Clouds
• 5.4.3 Conclusion
• 5.5 Heterogeneous computation of matrix products
• 5.5.1 A hybrid matrix multiplication application
• 5.5.1.1 The hardware and software used
• 5.5.1.2 Implementation option
• 5.5.1.3 Energy consumption
• 5.5.2 Experimental results of the matrix multiplication application
• 5.5.3 Conclusions
• 5.6 Summary

Inspec keywords: energy conservation; quality of service; telecommunication power management; power consumption; power aware computing

Other keywords: computation power; large scale equipment; communication-data sharing power; energy consumption; quality of service; energy aware ultrascale systems; quality metrics; multilevel hardware optimization; ultrascale infrastructures; energy awareness; energy efficiency; energy complexity; data access power; software optimizations

Subjects: Communication network design, planning and routing; Radio links and equipment; Energy conservation; Telecommunication systems (energy utilisation)