Data crunching is hard and more so when you have obscene amounts of data that are produced everyday and that you need to make sense of.
Most badass data crunchers are monstrosities called supercomputers which occupy entire rooms. They are basically your laptop on crack laced with speed with all the power of one of those decommissioned space shuttles.
The data intensive operations like climate prediction, cancer research etc as well as computationally heavy operations such as image analysis and graphics rendering require a hefty amount of data crunching capacity that is terribly expensive to firms whose owners do not own yachts.
Enter BOINC (Berkeley Open Infrastructure for Network Computing) An initiative of the University of California Berkley on exploring distributed computing of task. It started off as a project in support of the search for extraterrestrial life (Aliens basically) but now its used in climate research, protein synthesis, malaria control and a plethora of other applications.
Basically you get a desktop client which downloads some data to be crunched and your pc crunches the data in place of the screensaver and once done sends back the results/ a report to the server which issues more work if available. You can see how the millions of computers worldwide connected to the internet just sitting there running Microsoft Word and Outlook can be used for scientific betterment of humanity.
This is a purely volunteer affair where you specify how much of your processor is used, when to download and all manner of metrics to adhere to.
Intel has been pushing for its own brand of the same platform (seeing as they make the most processors) calling it progress thru processors.
I run this app though at minimal settings and I have done quite a sizeable amount of computing calculated in Ghrs.
Try it out.