Communications of the ACM
Smart Scheduling for Big Computing Tasks Cuts Emissions Up to a Third
Because the results of many large jobs are not needed immediately, there is scope to reduce carbon emissions with careful planning.
Credit: Alex Plavevski/EPA-EFE/Shutterstock
Researchers at Germany's Technical University of Berlin (TUB) found that shifting large computing jobs to nights or weekends can lower carbon emissions from datacenters by as much as a third.
They also recommend scheduling the jobs at times when renewable sources account for the greatest portion of the energy mix.
The researchers performed simulations for running data-intensive computing during a 16-hour window, and compared their results to a baseline test in which tasks were run each night at 1 a.m.
They found that running data-intensive computing during that window would reduce annual carbon emissions by 7.4% in Great Britain, 11.2% in Germany, and more than 33% in California.
TUB’s Philipp Wiesner said government policies could incentivize a similar system to reduce carbon emissions.
From New Scientist
View Full Article - May Require Paid Subscription
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA
No entries found