S a I S T Bo S T O N Mapping Parallel Iterative Algorithms onto Workstation Networks

For communication-intensive parallel applications, the maximum degree of concurrency achievable is limited by the communication throughput made available by the network. In previous work [HPS94], we showed experimentally that the performance of certain parallel applications running on a workstation network can be improved signi cantly if a congestion control protocol is used to enhance network performance. In this paper, we characterize and analyze the communication requirements of a large class of supercomputing applications that fall under the category of xed-point problems, amenable to solution by parallel iterative methods. This results in a set of interface and architectural features su cient for the e cient implementation of the applications over a large-scale distributed system. In particular, we propose a direct link between the application and network layer, supporting congestion control actions at both ends. This in turn enhances the system's responsiveness to network congestion, improving performance. Measurements are given showing the e cacy of our scheme to support large-scale parallel computations. This research was supported in part by NSF under grants IRI-8910195, IRI-9041581 and CDA-8920936.