Essay About Power Usage Effectiveness And Data Centers

Power Usage Effectiveness Case
PUE stands for power usage effectiveness. It is a measure of how effectively a company delivers power and cooling to the IT equipment. Ultimately, one of the goals of a company is to be more efficient in how they operate their data centers. Measuring PUE is a good place to start in considering how to reduce data center power consumption because the PUE ratio shows how much energy a company is using to power their equipment. The PUE is the ratio of total facility energy to IT equipment energy. The average PUE ratio is about 2.0, meaning for every watt of IT equipment energy used, two watts of total facility energy is used. The goal is to reduce overhead and get the PUE as close to 1.0 as possible. Thus, a company should aim to limit its total facility energy.The five methods recommended by Google for reducing power consumption are measuring power usage effectiveness, managing airflow, adjusting the thermostat, utilizing free cooling, and optimizing power distribution. Measuring power usage effectiveness helps measure how effectively a company is delivering power and cooling to their IT equipment. Managing airflow deals with eliminating the mixing of hot and cold air by using methods like computational fluid dynamics analysis and allowing data centers to receive the right amount of airflow to run effectively. Adjusting the thermostat means increasing the temperature in the data centers to about 80 degrees, which is the recommended temperature for data centers to operate safely. Utilizing free cooling involves using ambient temperatures outside of the data center to naturally cool equipment rather than using energy to cool. Finally, a company can optimize power distribution by minimizing conversion stages, thus saving money and power. Two percent of the world’s greenhouse gasses are a result of computing and of that two percent, fifteen percent of those emissions come from data centers. Data center power consumption is accelerating due to the world becoming more dependent on mobile devices like smart phones and tablets to obtain information. Cloud computing is also expected to accelerate power consumption. With the use of these devices and methods, data centers have to work more and that means companies have to use more energy to power them. However, more and more companies are attempting to reduce the energy they use through design choices and by methods like virtualization. Companies have an incentive not only reduce greenhouse emissions to help the planet but they can also help their wallets by reducing their costs by becoming more efficient. But with the competition within industries, some companies are not willing to share their methods for becoming more efficient so that they can get a leg up on other companies. I believe that data center power consumption is greatest within already developed economies. With economies that are already developed, there is a lot more competition than economies that are still developing. More competition means that more companies are operating and thus more energy is being used by all of these companies and industries. Also, these companies already have processes and have operated in some way for many years. These companies are probably less likely to want to change the way they operate so that they can reduce their power consumption. Some industries in already developed economies depend on the use of energy and it would be very difficult for them to change their ways. Developed economies are also more likely to possess the technology for power consumption whereas developing economies probably do not possess that technology yet.