lee@paul.rutgers.edu (Yong-fong Lee) (12/21/89)
I did not receive many responses to my query about the references
to the topic on compiling for distributed-memory machines. The
following is what I have collected. The contributors are listed
with their e-mail addresses, so anyone interested in their work
can issue his/her further query to the appropriate individuals.
Finally is some other references I know of and you may be interested
in.
Happy holidays to everyone.
Yong-fong (lee@cs.rutgers.edu)
---------------------------------------------------------------
From: Mike Quinn <quinn@umbra.cs.orst.edu>
Quinn, M. J., and Hatcher, P. J.,
Data Parallel Programming on Multicomputers, IEEE Software (to appear).
Quinn, M. J., and Hatcher, P. J.,
Compiling SIMD Programs for MIMD Architectures,
Proceedings of the IEEE Computer Society 1990 International
Conference on Computer Languages, March 1990, New Orleans (to appear).
Quinn, M. J., Hatcher, P. J., and Jourdenais, K. C.,
Compiling C* Programs for a Hypercube Multicomputer,
Proceedings of the ACM SIGPLAN Symposium on Parallel Programming:
Experience with Applications, Languages, and Systems,
July 1988, New Haven, pp. 57-65, SIGPLAN Notices 23, 9 (September 1988).
-------------------------------------------------------------------
From: eisen@verdix.com (Greg Eisenhauer)
You might check our work on Distributed Ada. I'm afraid I'm at another site
right now and I don't have all the references, but we've a paper in the
proceedings of Tri-Ada '89 and in IEEE Trans. on Soft. Eng. sometime around
last march... If you can't find anything, mail me next week at
eisen@src.honeywell.com and I can send you something.
-------------------------------------------------------------------
From: Henry.Thomas@irisa.fr (Henry Thomas -- TB 129 -- p 545)
Pandore
A System to Manage Data Distribution.
The goal of the Pandore project is to allow a user to execute on a parallel
machine programs originally written without explicit expression of
parallelism. The machines we consider are loosely-coupled, distributed
memory architectures.
Pandore allow the front-end user to ignore the low-level characteristics
of the distributed computer he is using, focusing on the logic data
organization to achieve parallelization. The user annotates data structures
to define a regular decomposition over a domain of virtual processors. This
domain corresponds to the ideal parallel architecture for the application.
Data may be partitioned or replicated over the domain.
-----------------------------------------------------------------
From: moe@dooma.prl.philips.nl (Petter Moe)
You might be interested in the machines built at the University of
Trondheim, see article:
Kjell Bratbergsengen:
"The development of HC16"
or something like that, in the proceedings of HCCA 4 (Fourth
Conference on Hypercubes, Concurrent Computers and Applications,
Monterey, spring '89.)
Also have a look at the DOOM approach, presented in proceedings from
PARLE '89, (Parallel Architectures and Languages Europe), see Lecture
Notes in Computer Science (Springer-Verlag) no. 365 & 366, article
W. Bronnenberg: "POOL and DOOM".
----------------------------------------------------------------
From: pstseng@flash.bellcore.com (Ping Sheng Tseng)
@PhDThesis(TsengAParallelizing,
Author="Tseng, P. S.",
Title="{A Parallelizing Compiler for Distributed Memory Parallel
Computers}",
School="Carnegie Mellon University ",
Month="May", Year="1989")
Also available in CMU CS TR
CMU-CS-89-148
------------------------------------------------------------------
From: chk@cs.purdue.edu (Chuck Koelbel)
We've just gotten a paper accepted at PPoPP next year. For the
moment, it exists as a TR (BiBTeX format):
@TechReport{blaze-tr89a,
author = "C. Koelbel and P. Mehrotra",
title = "Supporting Shared Data Structures on Distributed
Memory Architectures",
institution = "Purdue University",
year = "1989",
number = "CSD-TR 915",
address = "West Lafayette, IN",
month = oct
}
----------------------------------------------------------------------
Other references:
@Article(BePaDataOrgan,
Author = ``Bell and Patterson'',
Title = ``Data Organization in Large Numerical Computations'',
Journal = ``The Journal of Supercomputing'',
Year = 1987,
Volumn = 1,
Pages = 105--136 )
@Article(CaKeComProg,
Author = ``Callahan and Kennedy'',
Title = ``Compiling Programs for Distributed-memory Multiprocessors'',
Journal = ``The Journal of Supercomputing'',
Year = 1988,
Volumn = 2,
Pages = 151--170 )
@Article(CCLComPara,
Author = ``Chen and Choo and Li'',
Title = ``Compiling Parallel Programs by Optimizing Performance'',
Journal = ``The Journal of Supercomputing'',
Year = 1988,
Volumn = 2,
Pages = 171--208 )
@Inproceedings(KMRSemiAuto,
Author = ``Koelbel and Mehrotra and Rosendale'',
Title = ``Semi-automatic domain decomposition in BLAZE'',
Booktitle = ``ICPP'',
Year = 1987,
Pages = 521--524 )
@Article(ZBGSuperb,
Author = ``Zima and Bast and Gerndt'',
Title = ``{SUPERB: A tool for semi-automatic MIMD/SIMD
parallelization}'',
Journal = ``Parallel Computing'',
Year = 1988,
Volumn = 6,
Pages = 1--18 )
-----------------------------------------------------------------
END of report|
--------------