A framework for efficient execution of array-based languages on SIMD computers

Author

Prins, Jan F.

Author_Institution

Dept. of Comput. Sci., North Carolina Univ., Chapel Hill, NC, USA

fYear

1990

fDate

8-10 Oct 1990

Firstpage

462

Lastpage

470

Abstract

The author presents a framework for supporting efficient execution of machine-independent, array-based, data-parallel languages, such as Fortran-90 and Parallel Pascal, on distributed-memory SIMD (single-instruction-stream, multiple-data-stream) machines with mesh or hypercube interconnection topologies. The framework supports (1) a wide class of mappings of arrays into machines, (2) the implementation of many data selection and reorganization operations by manipulation of data descriptors instead of data movement, and (3) the decomposition of required data motions into sequences of efficient nearest-neighbor communications on the mesh. Each of these is discussed, and an application example is given. Related work is examined

Keywords

memory architecture; parallel architectures; parallel languages; Fortran-90; Parallel Pascal; array-based; data descriptors; data-parallel languages; distributed-memory SIMD; framework; interconnection topologies; machine-independent; nearest-neighbor communications; Computer languages; Computer science; Digital audio players; Hardware; Hypercubes; LAN interconnection; Parallel processing; Programming profession; Shape; Topology;

fLanguage

English

Publisher

ieee

Conference_Titel

Frontiers of Massively Parallel Computation, 1990. Proceedings., 3rd Symposium on the

Conference_Location

College Park, MD

Print_ISBN

0-8186-2053-6

Type

conf

DOI

10.1109/FMPC.1990.89497

Filename

89497