Home Environment Climate Change

Delaying Internet's Data Could Reduce Its Carbon Emissions

43
0

The servers, routers, switches and other components of a network are designed to work with much larger quantities of data than they do daily, but they use just about the same amount of electricity whether they’re idle or busy.

Sergiu Nedevschi of the University of California in Berkeley, US, and his colleagues from Intel Research labs in Berkeley and in Seattle, found out a way to save about 50% of the energy that a network consumes , by delaying data flowing into it by just a few milliseconds, during which the network adapters and servers could pass into sleep mode, saving energy.

For networks that are less used, this “sleeping” technique is most efficient at night, when fewer data is transferred. Microsoft researchers from Redmont, on the other hand, found out a problem that arises: that users working at night, by reserving just a single live connection, they are keeping up the entire server. That’s the worst case scenario. The problem is resolved by routing an incoming request to an already busy server, and keeping the other ones to a sleep mode.

The sleeping approach is practical, but would not make a big difference in energy use in the United States, because “routers and switches consume a very tiny amount of the annual energy here,” says Suresh Singh of Portland State University in Oregon, US.

“However, the impact could be significant in developing countries like India,” Singh adds, where the networks are growing and the electricity supplies are small.

Microsoft tested this theory by applying it to its Windows Live Messenger, and found out they got a 30% reduction in energy consumption during the 45 days of testing.

Sergiu Nedevschi’s /Intel and the Microsoft research projects were presented in papers at the USENIX Networked Systems Design and Implementation conference.

[via]

(Visited 50 times, 1 visits today)

LEAVE A REPLY

Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.