SETI: The People's Super Computer by Mark Stone

Previous: The Inefficiencies of the PC Revolution

"SETI" stands for the Search for Extra-Terrestrial Intelligence, a global endeavor managed by the SETI Institute. The main activities are to listen and to filter. Listening basically means using radio telescopes to scan the heavens picking up signals. However, the universe is a noisy place. Stars, planets, comets, pulsars… basically everything up there is giving off some kind of signal, however, faint, all the time. Filtering means sorting through all those signals to find the kind of organized, regular, complex patterns that would be indicative of communication by intelligent life elsewhere in the universe.

Filtering is a huge task. It requires a super computer, and truthfully we've yet to create a computer big enough for the task. In other words, we can listen faster than we can filter.

The folks at SETI were among the first to think deeply about the idea that (a) most computers were idle most of the time, and (b) most computers were connected to the Internet. In 1999, in partnership with UC Berkeley, they created a piece of software called SETI@Home.

The software works like this: a master server, originally at Berkeley, keeps track of all the filtering assignments to be done. Volunteers then download a small bit of software that

  • Checks to see when the volunteer's computer is otherwise idle
  • Contacts the master server for a filtering assignment
  • Runs the assignment during idle time
  • Returns the results to the master server

At present over 5 million computers contribute to the SETI@Home project, making it one of the largest distributed super computing endeavors in the world.

Next: Virtualization: One Step Back and Two Steps Forward

Add / view comments.