summary on developing functional parallel program 
Author Message
 summary on developing functional parallel program


Last week, I post the news called "develop parallel functional program".
I want to express my thanks for those who have privided me suggestions
and given me a lot of encouragement. Now I list  some of them below and hope
that we can have a further discussion on this topic. I set about to make
a survey on this topic, and any information and suggestion are welcomed!
Thanks in advance,

Hu Zhenjiang

University of Tokyo

                         LIST OF RESPONSES

== response ==

I would be interested in hearing about any responses you receive,

I would really appreciate it.


- Bill O'Connell
  (Illinois Institue of Technology)

== response ==

SISAL is a functional language for numeric computation.  The
parallelizing compiler runs on both shared- and distributed-memory

The SISAL compiler and symbolic de{*filter*} may be obtained by anonymous
ftp from: in ~ftp/pub/sisal.

information about the SISAL project.

Brian Grant

== response ==

You can have a look at:

Functional programming for loosely-coupled multiprocessors / Paul Kelly. -
London : Pitman ; Cambridge, Mass : MIT Press, 1989. - 253 p. ; 25 cm. -
(Research monographs in parallel and distributed computing, ISSN 0953-7767)
Complete rewrite of author's thesis, 1987. - Lit.opg. - Index.        
ISBN 0-273-08804-1
ISBN 0-262-61057-4 pbk


== response ==

Dear Hu Zhenjiang,

I have the following paper:

  "Exploiting parallelism in functional languages: a paradigm-oriented
   to appear in Workshop of Abstract Models for Parallel Computers,
   Leeds, April 1993.

I can send you a copy if you give me your postal address,

Best regards,


Computer Science
University of Hull                       Tel : +44 (0)482 465744
Hull HU6 7RX (UK)                        Fax : +44 (0)482 466666

== response ==

The follwing papers might be of interest to you.
They are available by anonymous ftp at in
directory pub/hains/textes.

author= {G. Hains and C. Foisy},
title= {The Data-Parallel Categorical Abstract Machine},
booktitle= {{PARLE'93} (to appear)},
year= 1993,
editor= {},
number= {},
series= {Lecture Notes in Computer Science},
pages= {},
publisher= {Springer},
categorie= "c",
note= {Draft as publication no. 842 } # diro


author= {G. Hains and L. M. R. Mullin},
title= {Parallel Functional Programming with Arrays},
journal= {The Computer Journal},
year= 1993,
categorie= "a"


Best regards,
Gaetan Hains

== response ==

Organization: Iowa State University, Ames, IA

Dear Hu Zhenjiang,
        I wish I had information to provide you with but I don't.
I am, however, very interested in this question you pose.  I would
love to hear what you found out from your post.  

I hope to hear from you in the future.

Take care

== response ==

 Dear Hu Zhenjiang!

My collegues and I are developing the same project,
last our paper has been accepted to conference in ???, Canada (May 93):

          S.M.Abramov, A.I.Adamowich, I.A.Nesterov,
               S.P.Pimenov, Yu.V.Shevchuck

   "Autotransformation of evaluation network as a basis
             for automatic dinamic parallelizing".


The paper described a computation model designed to organize
parallel computing. The computation is represented as an autotransformation
of an evaluation network consisting of processes and data. This model
can be used as a basis for a programming system with automatic dynamic
parallelizing of programs. It is supposed that the source language of the
system should be similar in its syntax to the conventional languages C,
Pascal, fortran without side effects. Languages are imperative
inside functions and are functional  - outside functions.
The grains of parallelism are the function of the source language.
The programming system is intended for parallel equipment based of
Intel microprocessors, Inmos Transputers, and the Texas Instruments
parallel DSPs.

 Sergey P.Pimenov                       PSI AS of Russia

== response ==

You might want to try M. Broy's approach to distributed (and real time) systems

        Author={Broy, M.},
        oitle= {Towards a Design Methodology for Distributed Systems},
        Crossref= {NATO.ASI.F55},

        Title= {Constructive Methods in Computing Science.
        {NATO ASI} Series Vol. {F}55},
        Booktitle= {Constructive Methods in Computing Science.
        {NATO ASI} Series Vol. {F}55},
        Editor= {M. Broy},
        Publisher= sv,
        Address= {Berlin},
        Year= 1989}

For a short exposition and a critic by L.Lamport of this and two other (nonfunctional) formal specification methods c.f. Distributed Computing Vol.6(1), 1992.

Good Luck,

Norbert V"olker

== response ==

Try "ftp" and cd to /pub/pcn.  Pcn (Program Composition
Notation) is an excellent functional parallel language.


Sat, 07 Oct 1995 15:50:43 GMT  
 [ 1 post ] 

 Relevant Pages 

1. Develop parallel program by functional approach

2. Why are functional programs shorter/faster to develop?

3. Parallel Functional Programming book

4. PhD Studentship: Parallel Functional Programming

5. Parallel Functional Programming Bibliography (2nd ed)

6. Parallel Functional Programming Bibliography

7. The History of Parallel Functional Programming

8. parallel functional programming

9. Functional Programming and Communicating Processes (SUMMARY)

10. Are there any archives for functional programs ? (SUMMARY)

11. SUMMARY: Debugging functional programs

12. Summary: functional programming of graph algorithms (long)


Powered by phpBB® Forum Software