Parallel implementations of combinations of broadcast, reduction and scan

Author

Wedler, Christoph ; Lengauer, Christian

Author_Institution

Fakultat fur Math. und Inf., Passau Univ., Germany

fYear

1997

fDate

17-18 May 1997

Firstpage

108

Lastpage

119

Abstract

Broadcast, reduction and scan are popular functional skeletons which are used in distributed algorithms to distribute and gather data. The authors derive new parallel implementations of combinations of broadcast, reduction and scan via a tabular classification of linearly recursive functions. The trick in the derivation is to not simply combine the individual parallel implementations of broadcast, reduction and scan, but to transform these combinations to skeletons with a better performance. These skeletons are also linearly recursive

Keywords

distributed algorithms; functional programming; parallel programming; recursive functions; broadcast; data distribution; data gathering; distributed algorithms; functional programming; functional skeletons; linearly recursive functions; linearly recursive skeletons; parallel implementations; performance; reduction; scan; tabular classification; Broadcasting; Cost function; Data structures; Distributed algorithms; Electrostatic precipitators; Functional programming; Parallel algorithms; Pattern matching; Skeleton;

fLanguage

English

Publisher

ieee

Conference_Titel

Software Engineering for Parallel and Distributed Systems, 1997. Proceedings., Second International Workshop on

Conference_Location

Boston, MA

Print_ISBN

0-8186-8043-1

Type

conf

DOI

10.1109/PDSE.1997.596831

Filename

596831