Electrical Engineering
      and Computer Sciences

Electrical Engineering and Computer Sciences

COLLEGE OF ENGINEERING

UC Berkeley

Tarmac: A Language System Substrate Based on Mobile Memory

Steven E. Lucco and David P. Anderson

EECS Department
University of California, Berkeley
Technical Report No. UCB/CSD-89-525
November 1989

http://www.eecs.berkeley.edu/Pubs/TechRpts/1989/CSD-89-525.pdf

Tarmac is a language system substrate on which systems for distributed parallel programming can be built. Tarmac provides a model of shared global state called mobile memory. The basic unit of state in this model can be viewed as both 1) a block of memory that can be directly accessed by machine instructions, and 2) a logical entity with a globally unique name that may be efficiently located, copied and moved. To support higher-level synchronization models, the movement of a memory unit may optionally enable computations.

Mobile memory is more flexible than models such as distributed virtual memory, shared tuple space, or distributed objects. It avoids the limitations of fixed page size, fixed data placement policy, and type-system or language dependence. This flexibility allows Tarmac to support a wide range of parallel programming models efficiently.


BibTeX citation:

@techreport{Lucco:CSD-89-525,
    Author = {Lucco, Steven E. and Anderson, David P.},
    Title = {Tarmac: A Language System Substrate Based on Mobile Memory},
    Institution = {EECS Department, University of California, Berkeley},
    Year = {1989},
    Month = {Nov},
    URL = {http://www.eecs.berkeley.edu/Pubs/TechRpts/1989/5743.html},
    Number = {UCB/CSD-89-525},
    Abstract = {Tarmac is a language system substrate on which systems for distributed parallel programming can be built. Tarmac provides a model of shared global state called mobile memory. The basic unit of state in this model can be viewed as both 1) a block of memory that can be directly accessed by machine instructions, and 2) a logical entity with a globally unique name that may be efficiently located, copied and moved. To support higher-level synchronization models, the movement of a memory unit may optionally enable computations. <p>Mobile memory is more flexible than models such as distributed virtual memory, shared tuple space, or distributed objects. It avoids the limitations of fixed page size, fixed data placement policy, and type-system or language dependence. This flexibility allows Tarmac to support a wide range of parallel programming models efficiently.}
}

EndNote citation:

%0 Report
%A Lucco, Steven E.
%A Anderson, David P.
%T Tarmac: A Language System Substrate Based on Mobile Memory
%I EECS Department, University of California, Berkeley
%D 1989
%@ UCB/CSD-89-525
%U http://www.eecs.berkeley.edu/Pubs/TechRpts/1989/5743.html
%F Lucco:CSD-89-525