Lectures

From energy efficient networking to sustainable networking [slides]
Prof. Michela Meo, Politecnico di Torino

Abstract. Energy efficient networking has become a hot topic in the last few years. In this tutorial, I will first motivate the interest and importance of this topic by presenting some data about energy consumption of network devices. Data will allow to identify the network segments that are the most energy hungry. The main approaches proposed in the literature for reducing network energy consumption will then be presented with particular attention to wireless access networks. Finally, we will shift our attention from energy efficient networking to the new challenging topic of sustainable networking.

HPC and energy savings [slides]
Prof. Jean-Marc Pierson, Université Paul Sabatier, Toulouse

Abstract. HPC has continuously and traditionally been energy hungry. HPC scientists are concerned first by performances and HPC machines have little chance to be switched off or even simply slowed down just to save some power. In this tutorial, we will investigate how IT researchers are trying to find solutions that do not decrease performance while saving energy and we will discuss the future of HPC towards Exascale architecture. Interestingly, analysis of power consumption, power modeling and power reduction is gaining interest in the HPC community. Some concrete examples will be given at application, operating systems, batch schedulers, processors and networks levels.

Energy Efficiency and Performance Trade-off in Communication Networks [slides]
Dr. Tuan Anh Trinh, Budapest University of Technology and Economics

Abstract. Dealing with energy efficiency typically points at savings in one or the other way. Saving communication resources is typically associated with compromises regarding communication quality. On this background, the tutorial focuses on (and somehow question) needs for trading off between energy efficiency and Quality of Experience (QoE). To this end, we will address energy efficiency metrics in communication networks on the one hand, and identify the views of users upon energy consumption in the context of Quality of Experience on the other hand. We will illustrate possible solutions, tools and techniques at different network layers to improve energy efficiency in communications networks without compromising QoE. Case studies will be provided with respect to mobile wireless networks (in particular LTE networks) and data center networks.

Energy-efficiency in Cloud data centers: the case of eco4cloud [slides – first part] [slides – second part]
Dr. Carlo Mastroianni, ICAR-CNR & Eco4cloud Srl

Abstract. The themes of green Cloud computing and energy-efficient data centers are today very topical, and we are witnessing a huge and increasing amount of academic and industrial research on such areas. One important issue is the limited use of green energy, as carbon and nuclear power provide a large percentage of the electricity used in data centers. Another main issue is that data centers are not used efficiently as most servers exploit only a fraction of their computational power. A viable solution is to reduce energy consumption by consolidating applications (hosted by “virtual machines”) on as few servers as possible, so as to put the remaining servers in a sleep state. The lecture will give an overview on the mentioned issues and will introduce attendees to eco4cloud, an approach for the adaptive and dynamic assignment of virtual machines to servers. The eco4cloud software and the tools (simulation, mathematical models, analysis of real traces) used to assess the performances will be presented. A demo will be given to show how virtual machines are managed in real data centers and how eco4cloud consolidates the load by dynamically assigning virtual machines to physical servers.

Energy-aware mobile computing [slides]
Dr. Carmela Comito, Università della Calabria, Rende

Abstract. The increasing computing power of mobile devices has opened the way to perform always more complex computations onboard the devices, including analysis and processing of data, in many real-life mobile scenarios, such as body-health monitoring, vehicle control, and wireless security systems. As mobile devices are battery-power operated, energy efficiency is a key concern to enable effective and reliable computing over them. One viable solution is tackling the problem of energy capacity shortage by distributing the energy consumption among the available devices promoting the cooperation among them. The aim of the lecture is to introduce the audience into the main issues of energy-efficient applications over cooperative mobile architecture, focusing on three main aspects: 1) energy characterization of both mobile devices and computational tasks; 2) energy-aware clustering of mobile devices ; 3) energy- efficient task allocation over a clustered network of mobile devices.

Green Performance Indicators for energy-efficient services [slides]
Prof. Barbara Pernici, Politecnico di Milano

Abstract. The talk will first survey the different types of green performance indicators proposed in the literature. In the second part of the talk, the experience with Green Performance Indicators in the GAMES EU FP7 project will be illustrated (http://www.green-datacenters.eu/), focusing on monitoring and use of such indicators to improve energy efficiency. Finally, current research within the ECO2Clouds EU FP7 project (eco2clouds.eu) will be presented, focusing not only on energy efficiency, but also on the evaluation of the environmental impact of applications running on clouds.

The new data center, metrics and energy paradigm Dr. Jean-Michel
Rodriguez, IBM France

Abstract. High-density equipment such as blade servers, 1U servers, and multi-core high-end servers requires concentrated power and cooling resources. Data center operators and IT executives are often uncertain about the capability of their existing data center and whether a new data center must be built to support higher rack densities. Consequently, new concepts and solutions are emerging helping customers to better select and manage IT resources and at the end building their Green IT strategy. The objectives of the lecture are to present the main points that customers have to face in mixed-density data center environment (cooling, powering, heterogeneous IT environments, etc.), see which designs, solutions and technologies are capable to ease those pain points, present the architecture of the Data Center of the future and the existing metrics to evaluate it.