A systematic approach for designing pipelined data parallel algorithms

Lionel Ming-Shuan Ni; Chung-Ta King

doi:10.1109/FMPC.1988.47512

A systematic approach for designing pipelined data parallel algorithms

Lionel Ming-Shuan Ni, Chung-Ta King

Research output: Contribution to conference › Conference Paper › peer-review

Abstract

A systematic procedure for designing pipelined data-parallel algorithms that are suitable for execution on multicomputers is introduced. This procedure concentrates on grouping loops in the original program so as to reduce the number of communicating processors, control the granularity, and increase the degree of pipelining. The procedure starts with a nested-loop program, manipulates the dependencies between the loops, and groups related loops to obtain pipelined and data-parallel operations. Using this procedure, it is possible to parallelize a nested loop automatically.< >

Original language	English
Pages	653-656
DOIs	https://doi.org/10.1109/FMPC.1988.47512
Publication status	Published - 1988
Externally published	Yes
Event	Proceedings of the 2nd Symposium on the Frontiers of Massively Parallel Computation - Duration: 1 Jan 1988 → 1 Jan 1988

Conference

Conference	Proceedings of the 2nd Symposium on the Frontiers of Massively Parallel Computation
Period	1/01/88 → 1/01/88

ISBNs

['0818658924']

Access to Document

10.1109/FMPC.1988.47512

Cite this

@conference{7fd1284fd0e34d1abf3633a761a31c6e,

title = "A systematic approach for designing pipelined data parallel algorithms",

abstract = "A systematic procedure for designing pipelined data-parallel algorithms that are suitable for execution on multicomputers is introduced. This procedure concentrates on grouping loops in the original program so as to reduce the number of communicating processors, control the granularity, and increase the degree of pipelining. The procedure starts with a nested-loop program, manipulates the dependencies between the loops, and groups related loops to obtain pipelined and data-parallel operations. Using this procedure, it is possible to parallelize a nested loop automatically.< >",

author = "Ni, \{Lionel Ming-Shuan\} and Chung-Ta King",

year = "1988",

doi = "10.1109/FMPC.1988.47512",

language = "English",

pages = "653--656",

note = "Proceedings of the 2nd Symposium on the Frontiers of Massively Parallel Computation ; Conference date: 01-01-1988 Through 01-01-1988",

}

TY - CONF

T1 - A systematic approach for designing pipelined data parallel algorithms

AU - Ni, Lionel Ming-Shuan

AU - King, Chung-Ta

PY - 1988

Y1 - 1988

N2 - A systematic procedure for designing pipelined data-parallel algorithms that are suitable for execution on multicomputers is introduced. This procedure concentrates on grouping loops in the original program so as to reduce the number of communicating processors, control the granularity, and increase the degree of pipelining. The procedure starts with a nested-loop program, manipulates the dependencies between the loops, and groups related loops to obtain pipelined and data-parallel operations. Using this procedure, it is possible to parallelize a nested loop automatically.< >

AB - A systematic procedure for designing pipelined data-parallel algorithms that are suitable for execution on multicomputers is introduced. This procedure concentrates on grouping loops in the original program so as to reduce the number of communicating processors, control the granularity, and increase the degree of pipelining. The procedure starts with a nested-loop program, manipulates the dependencies between the loops, and groups related loops to obtain pipelined and data-parallel operations. Using this procedure, it is possible to parallelize a nested loop automatically.< >

UR - https://openalex.org/W2142764052

U2 - 10.1109/FMPC.1988.47512

DO - 10.1109/FMPC.1988.47512

M3 - Conference Paper

SP - 653

EP - 656

T2 - Proceedings of the 2nd Symposium on the Frontiers of Massively Parallel Computation

Y2 - 1 January 1988 through 1 January 1988

ER -

A systematic approach for designing pipelined data parallel algorithms

Abstract

Conference

ISBNs

Access to Document

Other files and links

Fingerprint

Cite this