Data-intensive computing has become a key characteristic of modern large-scale scientific and engineering applications. For example, applications in the US Department of Energy's Accelerated Strategic Computing Initiative typically generate data in the range of hundreds of gigabytes to hundreds of terabytes. By 2005, detectors at the Large Hadron Collider at CERN, the European Laboratory for Particle Physics, will be producing several petabytes of data each year. The problem ASCI and similar applications face is that, with existing technologies, data sets the applications generate are much too large for effective storage, manipulation, archiving, navigation, visualization, or understanding.