The world's biggest physics experiment, the Large Hadron particle collider that began running Wednesday in a 17-mile tunnel below the French-Swiss border, produces so much data that even the massive computing power at the European Organization for Nuclear Research can't sift through it all.

So the Geneva-based lab, known by its old French acronym CERN, is sharing that burden among dozens of computing centers around the world. The result is the LHC Grid, a network of 60,000 computers that will analyze what happens when protons are hurled at each other inside the collider.

Scientists will need the additional computing power to sift through the mountains of data produced when the colliders' four giant detectors — 10 times more accurate than any previous instruments — measure activity at the subatomic level.

"You can think of each experiment as a giant digital camera with around 150 million pixels taking snapshots 600 million times a second," explains CERN's Ian Bird, who leads the grid project.

Sophisticated filters discard all but the most interesting data, but that still leaves 15 petabytes to be analyzed each year — enough to fill 2 million DVDs. So it will get sent via high-speed lines to 11 top research institutions in Europe, North America and Asia, and from there to a wider network of 150 facilities where the information can be scrutinized by thousands of researchers — who could share in the glory of any discovery.

Distributed computing has been seen before, notably in the SETI(at)Home project that lets people contribute their PCs' spare time to the search for extraterrestrial life. But the vast scale of the CERN effort will be closely watched by scientists who expect grid computing to become even more widely used for research ranging from new drugs to more effective nuclear power.