• Cite Icon22
  • https://doi.org/10.1109/fmpc.1995.380449Copy DOI Icon

Work-efficient nested data-parallelism

  • Feb 6, 1995
  • D.w Palmer +2 more
Show More
  • Abstract
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

An apply-to-all construct is the key mechanism for expressing data-parallelism, but data-parallel programming languages like HPF and C* significantly restrict which operations can appear in the construct. Allowing arbitrary operations substantially simplifies the expression of irregular and nested data-parallel computations. The technique of flattening nested parallelism introduced by Blelloch, compiles data-parallel programs with unrestricted apply-to-all constructs into vector operations, and has achieved notable success, particularly with irregular data-parallel programs. However, these programs must be carefully constructed so that flattening them does not lead to suboptimal work complexity due to unnecessary replication in index operations. We present new flattening transformations that generate programs with correct work complexity. Because these transformations may introduce concurrent reads in parallel indexing, we developed a randomized indexing that reduces concurrent reads while maintaining work-efficiency. Experimental results show that the new rules and implementations significantly reduce memory usage and improve performance. >

Similar Papers
  • Conference Article
  • Citations2

Supporting data-level and processor-level parallelism in data-parallel programming languages

  • Jan 05, 1993
  • P.J Hatcher +1
  • Research Article
  • Citations5

An efficient algorithm for communication set generation of data parallel programs with block-cyclic distribution

  • Apr 01, 2004
  • Parallel Computing
  • Gwan-Hwan Hwang
  • Research Article

An efficient algorithm for communication set generation of data parallel programs with block-cyclic distribution

  • Apr 01, 2004
  • Parallel Computing
  • G Hwang
  • Research Article
  • Citations1

Tuning Task Granularity and Data Locality of Data Parallel GPH Programs

  • Dec 01, 2001
  • Parallel Processing Letters
  • H Loidl
  • Book Chapter
  • Citations14

Data parallelism and functional programming

  • Jan 01, 1996
  • Björn Lisper
  • Book Chapter

Single Assignment C (SAC)

  • Jan 01, 2019
  • Clemens Grelck
  • Supplementary Content
  • Citations3

SIMD code generation in data-parallel programming

  • Jan 01, 2009
  • Publications of the UdS (Saarland University)
  • Nicolas Fritz
  • Conference Article
  • Citations42

A comparison of message passing and shared memory architectures for data parallel programs

  • Apr 01, 1994
  • Alexander C Klaiber +1
  • Research Article
  • Citations9

High Performance Fortran, Version 2

  • Dec 01, 1997
  • Parallel Processing Letters
  • Robert Schreiber
  • Conference Article
  • Citations2

Composites: trees for data parallel programming

  • Dec 17, 2002
  • M.C Carroll +1
  • Research Article
  • Citations9

A comparison of message passing and shared memory architectures for data parallel programs

  • Apr 01, 1994
  • ACM SIGARCH Computer Architecture News
  • A C Klaiber +1
  • Conference Article
  • Citations32

Size and access inference for data-parallel programs

  • Jan 01, 1991
  • Siddhartha Chatterjee +2
  • Research Article
  • Citations43

Asynchronous parallel simulation of parallel programs

  • May 01, 2000
  • IEEE Transactions on Software Engineering
  • S Prakash +2
  • Book Chapter

An Analysis of Programmer Productivity versus Performance for High Level Data Parallel Programming

  • Jan 01, 2011
  • Cole Alex +2
  • Conference Article
  • Citations5

SAC Goes Cluster: Fully Implicit Distributed Computing

  • May 01, 2019
  • Thomas Macht +1
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.