Issue No.04 - July/August (2001 vol.16)
Grid-based computing is needed to harness the large amounts of computing resources necessary to tackle data- and information-intensive problems. An exemplar and driving force has been the analyses required and anticipated in high-energy physics. Various very large numbers get tossed around; one realistic prediction is that, with the advent of a new generation of experiments, the CERN (European Center for Nuclear Research) facility will be generating several petabytes of data annually by 2005. For even the simplest sort of data analysis, this will require around 10 times more processing power per second than the most powerful supercomputer in existence. High-energy physics is not the only science demanding ever-greater computing power; astronomy, biological sciences, engineering design optimization, and climate modeling all guzzle CPU cycles and gobble storage.
Nigel Shadbolt, "Switching On to the Grid", IEEE Intelligent Systems, vol.16, no. 4, pp. 2-3, July/August 2001, doi:10.1109/MIS.2001.10011