Concurrent and distributed data structures for multikey sorting on computer clusters

Abstract

Summary form only given. This paper focuses on theoretical and practical aspects of the high-performance multikey sorting problem on computer clusters, with particular emphasis on the Alpha Maci Cluster, a world-class high-performance supercomputer that has many processors interconnected by a wide range of high-speed network connections. Even though the focus of this paper is on multikey sorting problems, developing new data structures and techniques for designing high-performance algorithms on computer clusters are of both theoretical and practical interest. We investigate strategies for developing, implementing, and refining high-performance algorithms for sorting multi-dimensional data on computer clusters. In addition, maximizing the performance of such distributed memory machines requires efficient data structures coupled with good load balancing.