Unstructured and dynamically varying algorithms are playing an increasingly important role in the solution of large-scale scientific problems on large-scale computers. This book focuses on the implementation of such algorithms on parallel computers, such as hypercubes and the Connection Machine®, that can be scaled up to incredible performances. The algorithms covered include those for partial differential equations and sparse linear algebra.
The nineteen contributions describe methods to effectively map fluids and structural mechanics codes that employ unstructured and/or adaptive meshes, scalable algorithms for problems in sparse linear algebra, scalable tools and compilers designed to handle irregular scientific computations, mapping methods for adaptive fast multipole methods, and parallelized grid generation and problem partitioning.
Piyush Mehrotra, Joel Saltz, and Robert Voigt are all on the staff at the Institute for Computer Applications in Science and Engineering (ICASE) located at the NASA Langley Research Center in Hampton, Virginia.
Piyush Mehrotra is a staff member at the Institute for Computer Applications in Science and Engineering (ICASE) located at the NASA Langley Research Center in Hampton, Virginia.
Joel Saltz is a staff member at the Institute for Computer Applications in Science and Engineering (ICASE) located at the NASA Langley Research Center in Hampton, Virginia.
Robert Voigt is a staff member at the Institute for Computer Applications in Science and Engineering (ICASE) located at the NASA Langley Research Center in Hampton, Virginia.
William Gropp is Director of the Parallel Computing Institute and Thomas M. Siebel Chair in Computer Science at the University of Illinois Urbana-Champaign.
Ewing Lusk is Argonne Distinguished Fellow Emeritus at Argonne National Laboratory.