Multinational collaborations on experiments at accelerators are common—detectors are paid for by the funding agencies of the people who build and use them—but the accelerators themselves are normally funded by their host country or region. In a departure from tradition, the CERN management invited nonmember states to participate financially and intellectually in the construction of the basic machine. "When the SSC died," Trilling says, "I was one of a group of people invited by the director of CERN to begin a discussion about potential involvement of the U.S."

Talks among representatives of the Department of Energy, the National Science Foundation, and CERN management began in 1995, soon after the LHC was formally approved by the CERN Council. Under terms approved by Congress and signed in December 1997 by Energy Secretary Federico Pe�a, National Science Foundation Director Neal Lane, CERN Council President Luciano Maiani, and CERN Director General Christopher Llewellyn Smith, the U.S. will invest $531 million in the accelerator and its two main detectors. About 85 percent of those funds will come from DOE and the remainder from NSF.

Others among the half dozen nonmembers who eventually signed on to help build the accelerator itself were Japan, Russia, and Canada. While the U.S. investment amounts to only ten percent of the LHC's cost including detectors, Americans will make up 20 percent of its users.

Berkeley Lab is a major player among the six national laboratories and five dozen American universities involved in the LHC. Through the Accelerator and Fusion Research Division, under the leadership of William Turner, the Lab is designing superconducting cable for the magnets in the main ring and will design and build cryogenics and quadrupole focusing magnets. And under the leadership of Gil Gilchriese of the Physics Division, the Lab is playing a crucial role in building the ATLAS detector.

ATLAS
Although the words A Toroidal LHC ApparatuS were jury-rigged to fit, the acronym ATLAS really just means big. Over 13 meters long, 22 meters in diameter, weighing 7,000 tons, ATLAS will be bigger than some five-story buildings. Yet all the action starts in the 5-centimeter-wide beam pipe running through its center. There, 40 million times a second, dense bunches of protons traveling in opposite directions will meet: 20 collisions at each crossing, 100 particles per collision, a shower of debris including gamma rays, electrons, muons, and—signatures of quarks—hadronic jets in which evidence of many other species is found. For a few microseconds each bit of jetsam that spews from these astronomically numerous events will leave evidence of its passing in different types of particle detectors using millions of integrated circuits distributed throughout the experiment. Their tracks will betray the presence of otherwise invisible entities such as massive supersymmetric "sparticles" and Higgs bosons.

"We can study only a small fraction of these events, so we have to sift the possibilities like a miner panning for gold," says Gil Gilchriese, who heads the Berkeley Lab group working on the ATLAS Inner Detector. "We establish electronic filters so that nearly one billion collisions every second are quickly culled to 100,000 of interest—the rest are dumped. A more sophisticated filter reduces that 100,000 to a few thousand, and finally a 'farm' of hundreds of computers picks a hundred or so of those to record on tape, so they can be studied later."