Computer Science
Energy Consumption of Computers
Energy consumption of computers refers to the amount of electricity used by computers and related equipment. This includes the energy used by the computer itself, as well as peripherals such as monitors, printers, and speakers. Reducing energy consumption can help to lower costs and reduce environmental impact.
Written by Perlego with AI-assistance
Related key terms
1 of 5
4 Key excerpts on "Energy Consumption of Computers"
eBook - PDF- Raoul-Abelin Choumin Nguemaleu, Lionel Montheu(Authors)
- 2014(Publication Date)
- Chapman and Hall/CRC(Publisher)
35 C H A P T E R 2 Green Computer Science Methodologies, Designs, Frameworks, and Tools That Can Be Used to Compute Energy Efficiently DEFINITIONS AND ABBREVIATIONS C: Programming language initially developed by Dennis Ritchie C++: Intermediate level programming language CPU: Central processing unit GHz: 1.000.000.000 Hz Hz: SI unit for frequency I/O: Input/output HIFI: High fidelity RAM: Random access memory ROM: Read only memory SI: International System of Units (abbreviated SI from French: Le Système national d’unités) 2.1 INTRODUCTION Nowadays, almost every system including mobile devices, HIFI equip-ment, data centers, wireless communication systems, and software sys-tems that run on computers or other systems need energy to work and are managed using software. That is why energy today is a very important factor in the design and development of software. Computer science that 36 ◾ Roadmap to Greener Computing better understands the properties of programs used to manage software will go a long way in reducing energy consumption. Energy efficient com-puting is currently used today in several domains. These include: • The development of large data centers for supporting web-based ser-vices such as search engines, email, online shopping, etc. • The development of mobile and embedded devices (the extension of battery time is a big challenge in this case). • The development of wireless communication. • The development of computer games. In this chapter, we analyze methodologies, concepts, and tools in com-puter science that can be used to improve energy efficiency in software. 2.2 ABOUT COMPUTER SCIENCE Computer science is the study of computers and computational systems. This includes theory, design, development, and application. Principal areas in computer science are artificial intelligence, computer systems, database systems, human factors, numerical analysis, programming languages, software engineering, and the theory of computing.
- Ishfaq Ahmad, Sanjay Ranka, Ishfaq Ahmad, Sanjay Ranka(Authors)
- 2016(Publication Date)
- Chapman and Hall/CRC(Publisher)
The trend of personal computers, portable computers and mobile devices will require long-lasting batteries, otherwise the user experience will be bad. To extend battery lifetime, we not only need to develop new battery techniques, but to improve the energy efficiency of these devices. Improving energy efficiency of mobile devices requires both hardware and software methods. The hardware methods aim to design low-power circuits and support dynamic power management strategies, and software methods try to make power/performance trade-offs with the APIs supplied by hardware. Finally, accompanied by the emerging of the Internet, many large data centers are built by big companies to supply stable, good quality service for customers. Even when the workload is low, these 649 650 Handbook of Energy-Aware and Green Computing servers still have very large power dissipation, because they are designed for the peak workload. These servers always need to be kept “on,” even though most of the time the workload is trivial. Furthermore, a similar amount of money, as used for energy consumption, needs to be spent on cooling down the data center. These reasons cause unproportional energy problems in data centers [4]. 28.1.1 Terminologies Before describing more details about power measuring and profiling, we first introduce some basic terminologies. • Energy : In computer systems, energy, in joules, is the electricity resource that can power hardware devices to do computation. More energy is used for the research of mobile platform and data centers. For mobile devices, energy is strongly related to battery lifetime. For data centers, which consume a large amount of energy, energy is used as the concern of electricity costs. Usually, researches in these areas use energy efficiency, such as PUE (power usage effectiveness), as the metric to evaluate their work. • Power : Power is the dissipation rate of energy. The unit of power is watt (W) (or joule per second).
- Carla Schlatter Ellis(Author)
- 2022(Publication Date)
- Springer(Publisher)
5 C H A P T E R 2 System Energy Models and Metrics This chapter presents basic energy terminology and widely accepted metrics. It describes mech- anisms provided by hardware to allow systems to manage how power is consumed in mobile devices. It addresses the measurement techniques for calibrating system models and tools for experimental evaluation. 2.1 MODELS OF HOW POWER/ENERGY IS CONSUMED IN MOBILE DEVICES In order to advance the state of the art in power/energy management, it is important to understand how power is currently used in mobile computing devices. This leads to building models of power consumption as a first step. 2.1.1 Power and Energy Models One method of characterizing the power consumption of a computing platform is to identify its major hardware components and determine how much of the overall power budget each component requires. This is often presented in the form of a pie chart, as shown in Fig. 2.1. This kind of graph can be useful in focusing upon the major consumers within a device. However, it is important to understand how such a chart has been generated, what data have been included, and what assumptions have been made about the workload of the device. Unfortunately, such graphs often appear without being accompanied by this information that is crucial to correct interpretations. For example, Fig. 2.1 suggests that, after the CPU, the DVD is the next most significant power hog in this laptop computer. Depending on the goals of the project, this conclusion may not lead in a productive direction. For example, if the goal is to extend battery lifetime and the platform is to be used primarily for accessing email and the web, then the potentially high power consumption of the DVD is never seen in practice. This motivates our discussion of the uses of the terms power and energy. This particular graph in Fig. 2.1 represents the maximum power consumption of each component in a Thinkpad R40 laptop, as described in Mahesri and Vardhan (2004).
eBook - ePubHarnessing Green IT
Principles and Practices
- San Murugesan, G. R. Gangadharan, San Murugesan, G. R. Gangadharan(Authors)
- 2012(Publication Date)
- Wiley-IEEE Computer Society Pr(Publisher)
Nowadays less than a full day's use of one's smartphone is intolerable to most. Notebook computer manufacturers are still striving for the elusive all-day PC, though some of the thin and light models are getting close. A big concern for data centres is the cost of electricity to run the servers and air-conditioning units to keep the room and servers cool. To make progress in improving energy efficiency, it is important to understand how energy is consumed. Figure 3.1 shows a typical power breakdown for a notebook PC during various operating conditions (provided by Si Software's Sandra benchmark suite). There are many components to target for improvement, and we believe that a focus on software will lead to significant impact. Figure 3.1 Breakdown of power usage on a notebook PC. 3.1.1 Processor Power States For software developers, the component of primary interest is the central processing unit (CPU). CPUs have defined energy states as well. It makes sense that if the CPU is not actively processing information or performing computations, it should be consuming minimal energy. The CPU has what are called C-states and P-states (Intel, 2011). C-states are core power states that define the degree to which the processor is ‘sleeping’. In state C0, the processor is active and executing instructions. Whilst in C0, the processor can operate at various frequency levels, designated as P-states (performance states). 3.1.1.1 C-States In the C0 state the CPU is active—it is busy carrying out some task, and it is performing that task at a frequency (P-state) that is appropriate. Between periods of activity the CPU can take the opportunity to rest or ‘sleep’. In fact, C-states are often referred to as sleep states. Intel processors support several levels of core and package (resources shared by all the cores) C-states that provide a flexible selection between power consumption and responsiveness
Index pages curate the most relevant extracts from our library of academic textbooks. They’ve been created using an in-house natural language model (NLM), each adding context and meaning to key research topics.
