It's about time we stopped wasting our idle time at ASU. Or maybe we'll start with our computers'.
Any computer's idle time could actually contribute to research on climate changes that may change the way we predict weather.
Scientists from Oxford University are leading a project in distributed computing to help analyze climate changes and use them to forecast weather fluctuations for this century. The experiment was developed for the British Broadcasting Corporation. The BBC kicked off the project in February with "Meltdown," a documentary on global warming.
To analyze climate changes, millions of measurements have to be made all over the world. It requires an enormous amount of supercomputing power to analyze all this information.
However, with the lack of supercomputers available on eBay, researchers have turned to another option. The idea behind distributed computing is to use thousands of people's computers with spare processing power to do calculations.
Most of the time, our computers don't process information to their full potential. For instance, my central processing unit usage is at about 7 percent right now.
There are many research projects that aim to utilize everyone's spare computing power for research that takes an enormous number of calculations.
All the computers being utilized in such a system send their results through the Internet to main computers that can compile all the data.
Scientists argue that using thousands of computers that already exist is a far more powerful and efficient method than using any supercomputer available.
The project works by running thousands of simulations of the world's different possible climate shifts and their effects. If participating in the project, you can view your own computer-generated simulation of global changes in a screensaver.
The purpose of these simulations is to assess how accurately we round our values and make predictions.
Several existing models for forecasting the weather raise questions about the effects of global warming. They all must approximate some values based on past temperature readings and evidence of changes in temperature in order to predict future changes.
The goal of this research project is to look at those approximations and at how minor changes that may have been ruled out as statistically insignificant may drastically change the entire outcome of the model.
Currently, there are only about 96,500 machines across the world participating in this particular research project. Most of the participants are in the United States and the United Kingdom.
The University of California, Berkeley, created Berkeley Open Infrastructure for Network Computing. BOINC is one of the largest organizing distributed computing infrastructures in the world. Its software has been used for various research projects, including this one.
There are thousands of computers in Phoenix alone. I'm not sure we have enough distributed computing research projects to go around for all the computers be involved. Maybe it's time ASU started a new center for using all our unused computing power.
Distributed computing projects like this one are popping up all over the place. Soon enough, I imagine people will be the only ones left wasting idle time.
Anjali is a molecular biosciences and biotechnology junior. Reach her at anjali.patel@asu.edu.
