MAR 01, 1993
Concurrent computation and high‐speed communication offer opportunities for simulations of more realistic physics.
Computers commonly process data and do simulations these days at rates of gigaflops ( floating‐point operations per second). In the next few years teraflop ( flops) computers will open an age in which we can study physics in new parameter domains, providing insights that may lead to new discoveries. These teraflop computers will be massively parallel, involving hundreds if not thousands of coupled processors tackling problems involving trillions of data points. To make effective use of these new resources will require new ideas about which problems should be solved, what numerical algorithms should be used to solve them, and what visualization and quantification methods should be used to enable scientists to comprehend and analyze the vast amounts of data that will be generated.
This article is only available in PDF format
© 1993 American Institute of Physics
