The authors gratefully acknowledge the help of all of the Concert project team, specifically Vijay Karamcheti, John Plevyak, and Xingbin Zhang, who built the Concert compiler and runtime system, and whose help enabled us to get our sequential and parallel execution times.
The research described in this paper was supported in part by NSF grants CCR-9209336 and MIP-92-23732, ONR grants N00014-92-J-1961 and N00014-93-1-1086 and NASA grant NAG 1-613. Andrew Chien is supported in part by NSF Young Investigator Award CCR-94-57809. Parallel machine runs used the Cray T3D machines at the Pittsburgh Supercomputing Center and the NASA Jet Propulsion Laboratory.