A nationwide parallel computing environment
Ken Kennedy, Charles F. Bender, John W. D. Connolly, John LeRoy Hennessy, Mary K. Vernon, Larry L. Smarr · Communications of the ACM · 1997
Since its beginning in 1985, the National Science Foundation's Supercomputer Centers Program has provided access to high-performance computing to the U.S. computational science and engineering community.Access initially meant classic shared memory parallel vector processor (PVP) computers, followed in 1989 by the first distributed massively parallel processing (MPP) machines.Access motivated the user community to an exponential increase in high-performance use that has continued to this day (see Figure 1). A Nationwide Parallel Computing EnvironmentHow to get there?Through a scalable software environment coupling the small parallelism (a few processors) of the desktop to the large-scale parallelism (thousands of processors) in Alliance machines.