A refinement of the code underlying the popular SETI@Home distributed computing application is now available under the name BOINC -- Berkeley Open Infrastructure for Network Computing. This will allow anyone (well, anyone with sufficient programming chops and an available server) to run distributed computing experiments. In effect, it will make supercomputing power available to the masses.

But BOINC is more than the server source code. It's also a client-side program, making it possible for a single user to participate in a variety of distributed computing projects with a single application. It's currently in beta.

This makes me really happy. Not only was SETI@Home the most well-known of the various distributed computing projects, it ran on a wide variety of platforms, including some not normally included in other projects, such as the Mac and SPARC. BOINC continues this tradition, running on Windows 95 and up, Linux/x86 (possibly others to come), Solaris/SPARC, and Mac OS X.

Once BOINC is out of beta, we could see a flowering of interesting distributed computing projects. What can be handled with this method? Anything that requires mathematical analysis of lots and lots of components. Current non-SETI distributed computing projects include protein folding and climate prediction.