Aliens No More

This month it was announced that SETI@home, perhaps the most famous volunteer mass computing project, is going into hibernation on the 31^st of March.  SETI@home was one of those rare projects that brought massively distributed computing into the public view.  It launched BOINC (pronounced “boink”) at UC Berkeley, and contributed enormously towards our ability to bring citizen-scientists, amateur technicians, and other members of the public into an easy-to-understand scenario where they could use their home computers to help with computational jobs that would simply be intractable for the organizations doing the research.

What was SETI@home?

The Search for Extraterrestrial Intelligence (SETI) was the brainchild of Frank Drake (best known for the Drake Equation.)  It monitored for radio transmissions from distant stars and searched to see if there was any pattern in them that might indicate it wasn’t just naturally occurring static, but rather communications by an extrasolar intelligence.  While collecting that radio data was a job for scientists with enormous radio telescopes, processing it to find these patterns was another thing all together.  In 1999, the University of California at Berkeley launched SETI@home, which was a screensaver that a user could install on their computers.  When the computer was otherwise idle, it would connect to servers at UC Berkeley, and download a packet of transmissions that needed to be analyzed, and then would do whatever analysis needed to be done, and would upload its results back to the university.

The project was a hit.  Thousands upon thousands of people downloaded the screensaver, and some would get whole groups of computers together and dedicated to helping.  SETI always engendered an interest in a certain segment of the population, and giving them a real, tangible way to contribute to the real science was extremely exciting for many. 

Why is it closing?

After over twenty years of operations, SETI@home announced its “hibernation.” This is happening for two major reasons: the diminishing returns of processing new data, and the workload for the staff in maintaining the project.  SETI@home doesn’t receive any government funding, so it faces many of the fund-raising challenges any non-profit does.  Although the project has clearly used a lot of hardware run-to-failure (as hardware failure and database outages have been cited as causes of downtime), often in those cases you’re just trading hardware costs for labor costs.  The practical upshot is that SETI@home might simply not be producing sufficient results to justify continuing it at this stage.

The SETI@home legacy

In 2002, the SETI@home project spawned the more general Berkeley Open Infrastructure for Network Computing (BOINC), which has built on the framework created for SETI@home to allow home users to contribute to a wide range of applications, as well as making important advancements in more commercial areas of distributed computing.  As of today, BOINC has a couple dozen different projects you can contribute computation time to, from Mathematics and astronomy to seismology and climatology.

These days most distributed computing you hear about in the news is about huge farms of graphics cards trying to mine BitCoins or other cryptocurrencies.  It’s nice to remember, though, that there are still thousands upon thousand of people globally that are spending their spare CPU cycles helping to advance science.