Txchnologist, NOVEMBER 7TH, 2011, BY MORGEN E. PECK

Think cloud. Then Now think rain. Let’s Now think spring. More Now let’s think green.On November 7th, 2011, Txchnologist’s Morgen E. Peck discussed the lack of clarity regarding the environmental sustainability of cloud computing. The term may suggest a state of ecological equilibrium. However cloud computing’s unpredictability and variability. For they suggest that greenness is contingent on how it is utilized. Cloud computing is intended to be highly efficient. Thereby allowing companies to outsource their computing and storage needs. Especially to providers like Google and Microsoft. Nevertheless, there is still much wonder and debate on how eco-friendly cloud computing actually is.

Cloud computing for the GM battery engineering teams.
GM battery engineering teams have tested and validated the A123 battery chemistry at the automaker’s Global Battery Systems lab in Warren, Mich. Teams from both companies will now work on developing calibrations and software controls for the battery system in preparation for production.

These massive investments in efficiency will result in data center energy consumption being cut. I mean by one-third by 2020. All according to the green tech analyst firm Pike Research.

New road, more cars

According to Pike Research, huge investments aimed at increasing efficiency. For it will result in data centers. Thereby consuming one-third less energy by 2020. Reuven Cohen, the founder of Enomaly, a Toronto-based company that creates products for cloud service providers. For he states that what was once stored on one server is now available in a multi-tenancy setting.

However, Jevons’ Paradox applies here. That’s providing more resources leads to more usage of those resources. Cohen also claims that cloud computing is likely more efficient, but it may persuade companies to take on more computing, which could eventually lead to greater power usage.

So Kerry Hinton, an electrical engineer at the University of Melbourne and part of the ARC Services Research Network. For Kerry believes that working out efficiency is an intricate process. Moreover, there is an important factor that is usually missed. Moving data from a private network to a consolidated data center. It is not an easy task. When a business moves all its info externally, its staff have to rely on the Internet each time. For they need access, and that consumes energy too. “The true issue in the long run is connecting to the internet. Then to get to the cloud provider,” Hinton states.

Not That Easy

Once a company moves all of that information offsite, its employees have to go through the Internet. I mean every time they need access. For those operations consume energy too. The data may be traveling on superhighways, but it has further to go. “The thing that’s going to kill us in the longer term is connecting into the internet to talk to the cloud provider,” says Hinton.

Energy hungry routers

According to Hinton, when an employee uses a cloud-based program instead of a desktop one to update a spreadsheet, a greater amount of energy is needed to establish an internet connection. Each time the spreadsheet is updated. All for the employee may be communicating with various data centers. All that span the globe. Data centers which require passing through 12-14 routers. Moreover, Hinton claims that routers are the most energy-consuming components of the internet. For they are consuming 7-10 nanojoules per bit.

Additionally, when customers use a wireless connection, the energy utilized increases significantly. That’s making data transportation a more prominent part of the efficiency equation. Hinton believes that cloud computing may be able to decrease energy usage, but this issue still needs to be addressed by providers, who often target new businesses.

Morgen E. Peck is a contributor to IEEE Spectrum, Innovation News Daily and other publications. Her last story for Txchnologist looked at “Skywalker” prosthetic technology.

This site is protected by wp-copyrightpro.com

%d bloggers like this: