Alien search merges with other home projects
SETI@home is getting a boost in computer power.
SETI@home, a downloadable screensaver that lets the public donate their unused computer time to the search for extraterrestrial intelligence, switches off today. But it is not going away: it is simply joining forces with similar distributed-computing projects on topics from climate models to cures for diseases. The move should boost the number of users, upping the computing power available to search for messages from alien life.
About a dozen projects are now signed up to a common software system, so that they can pool volunteers' computer time and use it more efficiently (see 'All for one'). As a result, each project should get access to more users, more of the time.
SETI@home, launched in May 1999, looks for regular or strong signals from outer space. It breaks up radio signals from the Arecibo radio telescope in Puerto Rico and sends these to the computers of 5.5 million volunteers, each of which analyses a small chunk of data and sends back the results.
But the project, says director David Anderson, has run out of steam and needs to take a new direction. "The science that SETI@home does is currently a dead end; it was meant to run for two years, it has now run for six. We are just rescanning the sky repeatedly and it's unlikely that we will find anything we haven't found before."
Scientific progress goes BOINC
The Berkeley Open Infrastructure for Network Computing (BOINC) platform will allow SETI@home to evolve, using data beamed directly from different telescopes, including ones in the Southern Hemisphere, and looking at a wider radiofrequency range. "We designed BOINC to let us do the things that we want to do in the future," Anderson says, including rolling out faster and more complex software.
Moving to the new infrastructure will not be difficult for Anderson. He is also director of BOINC, and his team built the software.
A computer scientist at the University of California, Berkeley, Space Science Laboratories, Anderson was driven to build BOINC because he realized that his and other stand-alone projects, such as Climateprediction.net, were also an inefficient use of resources. Users could subscribe to one or the other, but not both. When each of these projects hit a period of down time, their volunteer computing power went unused. By joining together, computing power can be used where it is needed most at any given time, boosting everyone's resources.
"We get a lot more participants this way," says Myles Allen, a physicist at the University of Oxford, UK, and founder of Climateprediction.net.
If you already have SETI@home or want to start contributing, you will have to download and install software from BOINC. You will then be given the option to donate various percentages of your machine's time to any or all of the projects. Users can chose, for example, to donate 70% of their unused computer power to finding curing disease, 20% to looking for aliens, and 10% to predicting the future climate.
Bigger is better
The list of BOINC projects is sure to get longer, its builders say, because the software makes it easy for scientists to use distributed computing in their pet projects.
Climateprediction.net took three large teams of programmers three years to build. "We and others were, to some extent, reinventing the wheel," says Tolu Aina, a programmer for the site. In contrast, adding something to BOINC is just a "student summer project", says François Grey, an technology expert at CERN, Europe's centre for particle physics.
Developing a single system also means the software is more reliable as more hands work to debug it, adds Allen. Many of the volunteers are programmers who, for fun, spot errors or possible improvements in the software and contribute solutions themselves. Stephen Pascoe, a climate researcher at Oxford's Rutherford Appleton Laboratory, remembers how volunteers once spotted that their climate model was accidentally generating "hailstones several metres across".
The boost in computing power may or may not bring SETI closer to its goal of making contact with intelligent life out there. But Anderson notes that BOINC brings much needed help to intelligent life on Earth, for solving problems closer to home.