Locality issues in a fine-grained parallel machine

https://doi.org/10.1016/0928-4869(95)00037-2Get rights and content

Abstract

This paper discusses mechanisms to exploit static information to increase locality in the Flagship fine-grained parallel machine. The machine supports a global address space across its distributed physical store. Load balancing is entirely dynamic and global-to-local address mapping to achieve high performance is carried out in a number of ways. A program is compiled into a graph of packets. This paper considers how static information is used to enable the dynamic mapping of the graph of packets to the distributed stores to improve performance and increase locality. Simulation figures suggest utilising static information results in some access taking less than half the time.

References (24)

  • G.S. Almasi et al.

    Highly Parallel Computing

    (1989)
  • M.P. Atkinson et al.

    Binding and type checking in database programming languages

    Comput. J.

    (1988)
  • P. Broughton et al.

    Designing system software for parallel declarative systems

    ICL Tech. J.

    (1987)
  • M.I. Greenberg

    An investigation into architectures for a parallel packet reduction machine

  • M.I. Greenberg, Personal communication,...
  • M.I. Greenberg et al.

    FLAGSHIP — A parallel reduction machine for declarative programming

    Comput. Control Engineering J.

    (1990)
  • S. Holdsworth et al.

    Aspects of protection on the Flagship machine: Binding, context and environment

    ICL Tech. J.

    (1989)
  • C.B. Jones

    Systematic Software Development Using VDM

    (1990)
  • J.A. Keane

    An overview of the Flagship system

    J. Functional Programming

    (1994)
  • J.A. Keane et al.

    A formal approach to parallel resource bindings

  • R.M. Keller et al.

    Applicative caching

    ACM TOPLAS

    (1986)
  • K.R. Mayes et al.

    Levels of atomic actions in the Flagship parallel system

    Concurrency: Practice Experience

    (1993)
  • Cited by (0)

    View full text