Bibliography on vectorizing compilers

thomas@emil.CSD.UU.SE (Thomas Lindgren)
Mon, 17 Sep 90 19:02:13 +0200

          From comp.compilers

Related articles
Bibliography on vectorizing compilers thomas@emil.CSD.UU.SE (1990-09-17)
| List of all articles for this month |

Newsgroups: comp.compilers
From: thomas@emil.CSD.UU.SE (Thomas Lindgren)
Keywords: vector, bibliography
Organization: Compilers Central
Date: Mon, 17 Sep 90 19:02:13 +0200

Here is the promised collected bibliography on vectorizing compilers.
Several formats (bibtex, refer and others) are used, but if I get the time I'll
convert them all into bibtex. Comments, suggestions and additions are
welcome.


Thomas Lindgren
thomas@emil.csd.uu.se
lindgren@sics.se


  Anyway, here is the complete list of ...
----------------------------8< cut here >8----------------------------
References on Vectorizing Compilers
======================================================================


This document is a collection of references on vectorizing compilers.


Contributors: (in alphabetic order)
James C Burley (burley@world.std.com)
Asha V. Chapnerkar (avc@cacs.usl.edu)
James Davies, Cray Comp. Corp. (jrbd@craycos.com)
Paul Havlak, Rice University (paco@rice.edu)
David Keppel, U. of Washington (pardo@cs.washington.edu)
Bron Campbell Nelson (bron@bronze.wpd.sgi.com)
Michael Wolfe (mwolfe@cse.ogi.edu)


(If you wish to have your organization properly attributed, please tell
me and it will be done.)


======================================================================
Reply-To: Paul Havlak <paco@rice.edu>


You may get other replies from Rice, but here goes anyway.
This is in bibtex format, hope you don't mind.


Paul


PFC (a project here at Rice)
============================


@Article{AlKe:Automatic,
                Author={J. R. Allen and K. Kennedy},
                Title={Automatic Translation of {FORTRAN} Programs to Vector Form},
                Journal=TOPLAS, Volume={9}, Number={4},
                Pages={491-542}, Month=Oct, Year={1987}}


@InProceedings{AlKe:Automatic-SIGPLAN84,
                Author={J. R. Allen and K. Kennedy},
                Title={Automatic loop interchange},
                BookTitle=SIGPLAN84,
                Month=Jun,
                Year={1984}}


@techreport{AlCK:Implementation,
author={ Randy Allen and David Callahan and Ken Kennedy},
title={ An Implementation of Interprocedural Analysis in a Vectorizing
Fortran Compiler },
year={1986},
month=May,
institution=COMPSCI,
number={TR86-38},
ignore={ }
}


@techreport{AlKe:Vector-86,
          author="Randy Allen and Ken Kennedy",
          title="Vector Register Allocation",
          institution="Dept. of Computer Science, Rice University",
          year=1986,
          month=Dec,
          number="COMP-TR 86-45"
}


IBM VS Fortran
==============
@Article{ScKo:86, Author={R.G. Scarborough and H.G. Kolsky},
                Title={A Vectorizing {FORTRAN} Compiler},
                Journal=IBMJRD, Volume={30}, Number={2},
                Month=Mar, Year={1986}}


Other compilers
===============
@TechReport{DoHi:Comparison,
                Author={J. J. Dongarra and A. Hinds]},
                Key={Vector Processing & CRAY & Vectorization & Compilers},
                Title={Comparison of the {CRAY X-MP-4}, the {F}ujitsu {VP-200},
                                and the {H}itachi {S-810/20}: an {A}rgonne perspective},
                Institution={Argonne National Laboratory}}


@TechReport{CaDL:Vectorizing,
                Author={D. Callahan and J. Dongarra and D. Levine},
                Title={Vectorizing Compilers: A Test Suite and Results},
                Institution={Argonne National Laboratory}, Number={109},
                Type={Technical Memorandum}, Month=mar, Year=1988,
                Note={To appear: {\it Proceedings of the
                                Supercomputing '88 Conference}}}


@InProceedings{AbMa:Vector, Author={W. Abu-Sufah and A. Malony},
                Title={Vector Processing on the Alliant {FX/8} Multiprocessors},
                BookTitle=ICPP86, Pages={559-566}, Month=Aug,
                Year=1986}




======================================================================
Reply-To: burley@world.std.com (James C Burley)


Try "Bulldog: A Compiler for VLIW Architectures", by John R. Ellis,
The MIT Press. It won the ACM Doctoral Dissertation Award 1985.


My previous company developed a Fortran optimizing compiler for their
128-bit-wide-instruction machine based on techniques described in an
early version of the thesis. It was quite a good compiler in many
ways, and its failings could have been addressed (from what I heard in
the halls) by implementing more of the techniques described in the
final version of the thesis published in the book.


                tq vm, (burley)


[ In fact, I did and found it very good. I second this. -- thomas]


======================================================================
Reply-To: bron@bronze.wpd.sgi.com (Bron Campbell Nelson)


Probably your best bet is to get Utpal Banerjee's new book:


Banerjee, Utpal. "Dependence Analysis for Supercomputing"
1988. Kluwer Academic Publishers, Norwell, Massachusetts.


since this book is recent and has a pretty good bibliography.


Two papers of particular note are Mike Wolfe's PhD dissertation:
Wolfe, M.J. "Optimizing Supercompilers for Supercomputers"
Oct. 1982. Ph.D. Dissertation, Report 82-1105,
Department of Computer Science, Univ. of Ill. at
Urbana-Champaign, Urbana, Illinois.


and Randy Allen's Ph.D. dissertation:
Allen, J.R. "Dependence Analysis for Subscripted Variables
and Its Application to Program Transformation"
Apr. 1983. Ph.D. Dissertation, Department of
Mathematical Sciences, Rice University, Hoston, Texas.


There are a number of other references, but these are 3 of the best,
and more can be found in Banerjee's, book.


--
Bron Campbell Nelson
bron@sgi.com or possibly ..!ames!sgi!bron


======================================================================
Reply-To: jrbd@craycos.com (James Davies)
Organization: Cray Computer Corporation


There are three papers on different variants of the KAP (the vectorizer
made by Dave Kuck's company) in the proceedings of the 1986 International
Conference on Parallel Processing. There is another paper on KAP in
the 1987 ICPP. ICPP tends to be full of reports on academic compiler
research as well.


The 1988 International Conference on Supercomputing (Boston) had a
session with 4-5 papers on vectorizing compilers at various hardware
companies. It included papers from Convex, Unisys, Stellar, and
Ardent and possibly others.


Jim Davies
jrbd@craycos.com




======================================================================
Reply-To: avc@cacs.usl.edu (Asha V. Chapnerkar)


I hope that these references help.


Asha V. Chapnerkar


References:


[Allen 82] Allen, John and Ken Kennedy, "PFC: A Program to Convert
Fortran to Parallel Form", Rice University, MaSc Technical
Report 82-6, 1982.


[Allen 82a] Allen, J.R, Ken Kennedy, Carrie Portfield, and Joe Warren,
"Conversion of Control Dependence to Data Dependence"
Technical Report, Department of Mathematical Sciences, Rice University,
1982.


[Cartwright 89] Cartwright, Robert and Matthias Felleisen, \fI
"The Semantics of Program Dependence"\fR,
Proceedings of SIGPLAN Symposium on Compiler Construction, 1989,
pp. 13-27.


[Furtney 82] Furtney, Mark, and Terrence Pratt, \fI"Kernel-Control Tailoring
of Sequential
Programs for Parallel Execution"\fR, Proceedings of the 1982 International
Conference on Parallel Processing, pp. 245-247.


[Jouvelot 87] Jouvelot, Pierre, \fI"Semantic Parallelization: A practical exercise in abstract
interpretation"\fR, Proceedings of the 14th Annual Symposium on Principles
of Programming Languages, 1987, pp. 39-48.


[Jouvelot 86] Jouvelot, Pierre, \fI"Parallelization by Semantic
Detection of Reductions"\fR,
LNCS 213: ESOP 86, Springer-Verlag, Berlin, 1986, pp. 223-236.


[Kuck 74] Kuck, David, Paul Budnick, Ross Towle, and et. al., \fI" Measurements
of Parallelism in Ordinary Fortran Programs"\fR, Computer, January 1974,
pp. 37-45.


[Kuck 81] Kuck, David, and et. al., \fI"Dependence Graphs and
Compiler Optimizations"\fR, Proceedings of the Eighth Annual ACM
Symposium on Principles of Programming Languages, 1981, pp. 207-218.


[Wolfe 82] Wolfe, Michael J., \fBOptimizing Supercompilers for
Supercomputers\fR, Ph. D. Thesis, Department of Computer Science,
University of Illinois at Urbana-Champaign, 1982.




======================================================================
Reply-To: Michael Wolfe <mwolfe@cse.ogi.edu>


References on vectorizing compilers, eh?


The Cohagan and Wedel papers are about the old compiler for the
TI-ASC, back in the late 60's and early 70's.


The Schneck paper is another very early vectorization paper, though
it may not have been implemented in a commercial product.


Five papers give info about commercial compilers from
Fujitsu, IBM, Unisys, Ardent, Convex


Two papers about research compilers (from France and Japan).


One paper about the Cydrome compiler, which is not strictly speaking
a vectorizing compiler.


+++


two old ones about the vectorizing compiler for the TI-ASC:


%A William L. Cohagan
%M Texas Instruments, Inc.
%T Vector Optimization for the ASC
%B Proc. of the Seventh Annual Princeton Conf. on Information Sciences and Systems
%P 169-174
%I Princeton University
%C Princeton, NJ
%D 1973
%Z March 22-23, 1973, Princeton Univ.
%K data dependence, vectorization, TI-ASC
%X
This paper describes the vectorization process used for the
Texas Instruments Advanced Scientific Computer (TI-ASC) compiler,
then two years old.
It describes the three-dimensional loop structure that the ASC
hardware can execute,
and introduces inter- and intra-statement "dependencies",
loop normalization (canonical loop form), and several rules for
deciding when a loop can be vectorized, including multiple nested loop
vectorization.
(mw)


%A Dorothy Wedel
%M Texas Instruments, Inc.
%T Fortran for the Texas Instruments ASC System
%J SIG\PLAN Notices
%V 10
%N 3
%P 119-132
%D March 1975
%K vectorization, interchanging, restructuring, TI-ASC
%X
This paper talks about the NX Fortran compiler for the Texas
Instruments Advanced Scientific Computer.
It describes some array language extensions, such as the SUBARRAY
statement and array cross sections.
Hazard analysis is used for "vectorization" (perhaps the first
use of this word), with temporary array allocation, loop interchanging
and loop reversal (running backwards) also attempted.
Use of multiple pipes is optimized by breaking vector operations
into multiple instructions or finding multiple independent vector
operations.
(mw)




Another old one about early work in vectorization; I don't know
if this was related to a specific compiler:


%A Paul B. Schneck
%M Goddard Institute for Space Studies
%T Automatic Recognition of Vector and Parallel Operations in a Higher Level Language
%J SIG\PLAN Notices
%V 7
%N 11
%P 45-52
%D November 1972
%Z Special Issue on Control Structures in Programming Languages
%K vectorization
%X
This is an early paper about an automatic Fortran vectorizer.
It uses interval analysis and strongly connected regions of the flow
graph and extended basic blocks.
It also shows vector output with "*" subscripts, expanded scalars,
type "A" IF unswitching and partial and subexpression vectorization.
(mw)




Recent papers about commercial compilers from
Fujitsu (early, 1983), IBM (1986) Unisys (1987), Ardent (1988), Convex (1988)


%A Sachio Kamiya
%A Fumio Isobe
%A Hidea Takashima
%A Masaaki Takiuchi
%% Practical Vectorization Techniques for the FACOM VP
%B Information Processing 83
%E R. E. A. Mason
%I Elsevier Science Publishers (North-Holland)
%D 1983
%P 389-394
%X
Fujitsu vectorizer
(mw)


%A Randolph G. Scarborough
%A Harwood G. Kolsky
%M IBM
%T A vectorizing Fortran compiler
%J IBM J. Res. Develop.
%V 30
%N 2
%D March 1986
%P 163-171
%K vectorization, IBM


%A Howard B. Coleman
%M Unisys
%T The Vectorizing Compiler for the Unisys ISP
%B Proc. of the 1987 International Conf. on Parallel Processing
%E Sartaj K. Sahni
%P 567-576
%I Penn State Press
%C University Park, PA
%D 1987
%O August 17-21, 1987
%K vectorization
%X
Basic vectorization in a compiler.
(mw)


%A Randy Allen
%M Ardent
%T Unifying Vectorization, Parallelization and Optimization: The Ardent Compiler
%B Proc. of the 1988 International Conf. on Supercomputing
%V II
%P 176-185
%D 1988
%K parallelization, vectorization
%X
Ardent compiler is wonderful.
(mw)


%A Randall Mercer
%T The Convex Fortran 5.0 Compiler
%B Proc. 1988 Intl Conf. on Supercomputing
%V 2
%P 164-175
%D 1988
%K Convex, vectorization




Two research compilers:


%A Babak Dehbonei
%A Gerard Memmi
%T Velour: A New Vectorizing Compiler Prototype
%B Proc. 1988 Intl Conf. on Supercomputing
%V 3
%P 431-440
%K vectorization


%A Takao Tsuda
%A Yoshitoshi Kunieda
%M Kyoto Univ.
%T V-Pascal: an Automatic Vectorizing Compiler for Pascal with No Language Extensions
%B Proc. of Supercomputing 88
%P 182-189
%I IEEE Computer Society Press
%C Los Angeles
%D 1988
%O Orlando, FL, November 14-18, 1988
%K vectorization
%X
Vectorization is done on nested loops in Pascal, using indirect
addressing lavishly.
Good performance results on Hitachi S-810 and S-820 systems.
(mw)




This compiler isn't a vectorizing compiler, strictly speaking, but
has many of the same characteristics and features:


%A James C. Dehnert
%A Peter Y.-T. Hsu
%A Joseph P. Bratt
%M Cydrome
%T Overlapped Loop Support in the Cydra 5
%B Proc. of Third Intl. Conf. on Architectural Support for Programming Languages and Operating Systems
%P 26-38
%I ACM Press
%C New York
%D April, 1989
%K Cydrome, data dependence, vectorization, VLIW
%X
The Cydra 5 was designed for overlapped loop support using wide word
instructions.
The compiler then rolls loops around themselves, as the 7600 family did,
to get parts of multiple iterations going all at the same time, but
packing the results into wide instruction words.
The architectural support for this approach is described as well as the
compiler algorithms.
(mw)




======================================================================
Reply-To: pardo@cs.washington.edu (David Keppel)


Here's a dump of our local guru's database. I know there's some
things missing, because I have copies. If you can wait, I'll be back
to the Northwest on the 25th; send me e-mail & I can give you names of
some papers, including one that compares features and code quality of
various compilers (the paper is valuable but I'm not convinced it does
a fair job). You'll have to sort through the following yourself.


;-D on ( Vectorizing netnews ) Pardo


51 Citations Follow:
%X /u1/rrh/bib/rrh.d/arch90.ref
%T VAX Vector Architecture
%A Dileep Bhandarkar
%A Richard Brunner
%J ISCA90
%K computer architecture comp arch


%X /u1/rrh/bib/rrh.d/asplos89.ref
%J ASPLOS89
%T A Unified Vector/Scalar Floating-Point Architecture
%A Norman Jouppi
%A Jonathan Bertoni
%A David W. Wall


%X /u1/rrh/bib/rrh.d/popl89.ref
%T The Program Dependence Graph and Vectorization
%A William Baxter
%A Henry R. Bauer
%P 1-11
%J POPL89
%K popl 1989 popl16
%K pdg


%X /u1/rrh/bib/rrh.d/sigplan88.ref
%J CCC88
%T Compiling C for Vectorization, Parallelization, and Inline Expansion
%A Randy Allen
%A Steven C. Johnson
%P 241-249


%X /u1/rrh/bib/rrh.d/cg.ref
%T A Retargetable Vector Code Generator
%A Tom C. Reyes
%J AFIPS CONF PROC, 1987 NCC
%D 1977
%P 657


%X /u1/rrh/bib/rrh.d/compiler.ref
%A Timothy A. Budd
%T An APL Compiler for a Vector Processor
%J TOPLAS
%V 6
%N 3
%D JUL 1984
%P 297-313


%X /u1/rrh/bib/rrh.d/compiler.ref
%A Timothy A. Budd
%T Time Space Tradeoffs in Vector Algorithms for APL Functions
%J SIGPLAN
%V 23
%N 12
%D DEC 1988
%P 63


%X /u1/rrh/bib/rrh.d/compiler.ref
%A R. E. Millstein
%A C.A. Muntz
%T The ILLIAC IV Fortran Compiler
%J Proc. of Conf. on Programming Languages and Compilers for Parallel and Vector Machines
%B SIGPLAN
%V 10
%N 3
%D MAR 1975
%P 1-8


%X /u1/rrh/bib/rrh.d/compiler.ref
%A D. L. Presberg
%A N. W. Johnson
%T The Paralyzer: IVTRAN's Parallelism Analyzer and Synthesizer
%J Proc. of Conf. on Programming Languages and Compilers for Parallel and Vector Machines
%B SIGPLAN
%V 10
%N 3
%D MAR 1975
%P 9-16


%X /u1/rrh/bib/rrh.d/compiler.ref
%T Program Optimization -- Theory and Practice
%A David B. Loveman
%A Ross A. Faneuf
%J SIGPLAN
%V 10
%N 3
%J PROC of a Conference on Programming Languages and Compilers for Parallel and Vector Machines
%D MAR 1975
%P 97-102


%X /u1/rrh/bib/rrh.d/peterd.ref
%T The Power of a One-Dimensional Vector of Processors
%A Jon Louis Bentley
%A Thomas Ottmann
%P 80
%J PROC INTL Workshop on Graphtheoretic Concepts in Computer Science (WG 80), (LNCS 100)
%D JUN 1980
%K parallel processing
%X describes efficient implementation of matrix multiply
%X and minimum spanning trees
%X -peterd 5/27/88


%X /u1/rrh/bib/rrh.d/peterd.ref
%T Automatic Translation of FORTRAN Programs to Vector Form
%A Randy Allen
%A Ken Kennedy
%J TOPLAS
%V 9
%N 4
%D OCT 1987
%P 491-542


%X /u1/rrh/bib/rrh.d/peterd.ref
%T Vectorization of Computer Programs with Applications to Computational Fluid Dynamics
%A Wolfgang Gentzsch
%J Notes on Numerical Fluid Mechanics
%V 8
%D 1984


%X /u1/rrh/bib/rrh.d/peterd.ref
%T A New Approach to Vector Code Generation for Applicative Languages
%A Timothy A. Budd
%R TR 88-60-18
%I DCS, Oregon State University
%C Corvallis, OR
%D AUG 1988


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Robert E. Millstein
%A Charles A. Muntz
%T The Illiac IV Fortran Compiler
%J PROC of a CONF on Programming Languages and Compilers for Parallel and Vector Machines
%J SIGPLAN
%V 10
%N 3
%D MAR 1975
%P 1-8


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Michael K. Donegan
%A Stuart W. Katzke
%T Lexical Analysis and Parsing for a Vector Machine
%J PROC of a CONF on Programming Languages and Compilers for Parallel and Vector Machines
%J SIGPLAN
%V 10
%N 3
%D MAR 1975
%P 138-145
%K parallel compiler


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Howard E. Krohn
%T A Parallel Approach to Code Generation
%J PROC of a CONF on Programming Languages and Compilers for Parallel and Vector Machines
%J SIGPLAN
%V 10
%N 3
%D MAR 1975
%P 146-152
%K parallel compiler


%X /u1/rrh/bib/rrh.d/parallel.ref
%A D. J. Kuck
%A R. H. Kuhn
%A B. Leasure
%A M. Wolfe
%T The Structure of an Advanced Vectorizer for Pipelined Processors
%J PROC IEEE 4th COMPSAC
%D 1980
%P 709-715
%X INTL Computer Software and Applications CONF


%X /u1/rrh/bib/rrh.d/parallel.ref
%A C. N. Arnold
%T Performance Evaluation of Three Automatic Vectorizer Packages
%J ICPP82
%D 1982
%P 235-242
%X have it
%X KAP (Kuck and Associates), VAST (Pacific Sierra Research), and
%X CDC Cyber 200 Fortran compared on Livermore kernels (18)


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Paul B. Schneck
%T Movement of Implicit Parallel and Vector Expressions Out of Program Loops
%J PROC of a CONF on Programming Languages and Compilers for Parallel and Vector Machines
%J SIGPLAN
%V 10
%N 3
%D MAR 1975
%P 103-106


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Ron Cytron
%T Doacross: Beyond Vectorization for Multiprocessors (Extended Abstract)
%J ICPP86
%D AUG 1986
%P 836-844


%X /u1/rrh/bib/rrh.d/parallel.ref
%A John R. Allen
%A Ken Kennedy
%T Automatic Translation of Fortran Programs to Vector Form
%R Rice Technical Report 476-029-4 (re-released as Rice COMP TR84-9, JUL 1984)
%I Rice University
%C Houston, TX
%D OCT 1980
%X Baer has it, now Rice COMP TR84-9, JUL 1984


%X /u1/rrh/bib/rrh.d/parallel.ref
%A C. Arnold
%T Vector Optimization on the Cyber 205
%J ICPP83
%D AUG 1983


%X /u1/rrh/bib/rrh.d/parallel.ref
%A James Davies
%A Christopher Huston
%A Thomas Macke
%A Bruce Leasure
%A Michael Wolfe
%T The KAP/S-1: An Advanced Source-to-Source Vectorizer for the S-1 Mark IIa Supercomputer
%J ICPP86
%D AUG 1986
%P 833-835


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Christopher Huston
%A Thomas Macke
%A James Davies
%A Michael Wolfe
%A Bruce Leasure
%T The KAP/205: An Advanced Source-to-Source Vectorizer for the Cyber 205 Supercomputer
%J ICPP86
%D AUG 1986
%P 827-832


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Dorothy Wedel
%T Fortran for the Texas Instruments ASC System
%J PROC of a CONF on Programming Languages and Compilers for Parallel and Vector Machines
%J SIGPLAN
%V 10
%N 3
%D MAR 1975
%P 119-132


%X /u1/rrh/bib/rrh.d/parallel.ref
%A S. Kamiya
%A F. Isobe
%A H. Takashima
%A M. Takiuchi
%T Practical Vectorization Techniques for the Facom VP
%J IFIP Congress PROC
%D 1983
%P 389-394


%X /u1/rrh/bib/rrh.d/parallel.ref
%A R. H. Perrott
%A D. Crookes
%A P. Milligan
%A W. R. M. Purdy
%T Implementation of an Array and Vector Processing Language
%J IEEE
%D 1983
%P 232-239


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Michiaki Yasumura
%A Yoshikazu Tanaka
%A Yasusi Kanada
%A Akio Aoyama
%T Compiling Algorithms and Techniques for the S-810 Vector Processor
%J IEEE
%D 1984
%P 285-290


%X /u1/rrh/bib/rrh.d/parallel.ref
%A L. Higbee
%T Vectorization and conversion of Fortran Programs for the Cray-1 CFT Compiler
%R Publication 2240207
%I Cray Research Inc.
%C Mendota Heights, MN
%D JUN 1979


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Alain Bossavit
%A Bertrand Meyer
%T An Application of Program Transformation to Supercomputer Programming
%P 27-38
%X development of a vectorized algorithm (in Ada) from a recursive algorithm
%B Vector and Parallel Processors in Computational Science
%E I. S. Duff
%E J. K. Reid
%D 1985
%I North-Holland
%C Amsterdam
%X PROC 2nd INTL CONF Vector and Parallel Processors in Computational Science
%X (VAPP) AUG 1984
%X Computer Physics Communications v37 (1985)


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Ron Cytron
%A David J. Kuck
%A Alex V. Veidenbaum
%T The Effect of Restructuring Compilers on Program Performance for High-Speed Computers
%P 39-48
%B Vector and Parallel Processors in Computational Science
%E I. S. Duff
%E J. K. Reid
%D 1985
%I North-Holland
%C Amsterdam
%X PROC 2nd INTL CONF Vector and Parallel Processors in Computational Science
%X (VAPP) AUG 1984
%X Computer Physics Communications v37 (1985)


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Toshihiko Matsuura
%A Kenichi Miura
%A Mitsuhiro Makino
%T Supervector Performance without Toil: Fortran Implemented Vector Algorithms on the VP-100/200
%P 101-107
%B Vector and Parallel Processors in Computational Science
%E I. S. Duff
%E J. K. Reid
%D 1985
%I North-Holland
%C Amsterdam
%X PROC 2nd INTL CONF Vector and Parallel Processors in Computational Science
%X (VAPP) AUG 1984
%X Computer Physics Communications v37 (1985)


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Peter L. Bird
%T Linear Time Detection of Inherent Parallelism in Sequential Programs
%P 69-76
%X flow analysis for structured programs, nothing new here
%B Vector and Parallel Processors in Computational Science
%E I. S. Duff
%E J. K. Reid
%D 1985
%I North-Holland
%C Amsterdam
%X PROC 2nd INTL CONF Vector and Parallel Processors in Computational Science
%X (VAPP) AUG 1984
%X Computer Physics Communications v37 (1985)


%X /u1/rrh/bib/rrh.d/parallel.ref
%A K. Y. Lee
%A W. Abu-Sufah
%A D. J. Kuck
%A On Modeling Performance Degradation Due to Data Movement in Vector Machines
%J ICPP84
%D AUG 1984
%P 269-277


%X /u1/rrh/bib/rrh.d/parallel.ref
%A D. A. Calahan
%A Influence of Task Granularity on Vector Multiprocessor Performance
%J ICPP84
%D AUG 1984
%P 278-284


%X /u1/rrh/bib/rrh.d/parallel.ref
%T An Optimal Instruction Scheduling Model for a Class of Vector Processors
%A S. Ayra
%J IEEETC
%V C34
%N 11
%D NOV 1985
%P 981-995
%X -peterd 4/1/87


%X /u1/rrh/bib/rrh.d/parallel.ref
%T Code Generation for Partially Vectorizable Loops in the Vectorizing Pascal-XT Compiler
%A C. Hammer
%A G. Raebel
%P 295-302
%J CONPAR 86, CONF on Algorithms and Hardware for Parallel Processing (LNCS 237)
%D SEP 1986


%X /u1/rrh/bib/rrh.d/parallel.ref
%T Automatic Vectorization for High Level Languages Based on an Expert System
%A Thomas Brandes
%P 303-310
%J CONPAR 86, CONF on Algorithms and Hardware for Parallel Processing (LNCS 237)
%D SEP 1986


%X /u1/rrh/bib/rrh.d/parallel.ref
%T Vector Processing on the Alliant FX/8 Multiprocessor
%A W. Abu-Sufah
%A A. D. Malony
%P 559-566
%J ICPP86
%D AUG 1986


%X /u1/rrh/bib/rrh.d/parallel.ref
%T Programming in VS Fortran on the IBM 3090 for Maximum Vector Performance
%A Bowen Liu
%A Nelson Strother
%J COMPUTER
%V 21
%N 6
%D JUN 1988
%P 65


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Randy Allen
%T Unifying Vectorization, Parallelization, and Optimization: The
Ardent Compiler
%J PROC Supercomputing '88: Technology Assessment, Industrial
Supercomputer Outlooks, European Supercomputing Accomplishments, and
Performance & Computations
%I Third INTL CONF on Supercomputing (ICS '88)
%C St. Petersberg, FL
%V 2
%D 1988
%P 176-185


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Howard B. Coleman
%T The Vectorizing Compiler for the Unisys ISP
%J PROC Supercomputing '88: Technology Assessment, Industrial
Supercomputer Outlooks, European Supercomputing Accomplishments, and
Performance & Computations
%I Third INTL CONF on Supercomputing (ICS '88)
%C St. Petersberg, FL
%V 2
%D 1988
%P 186-195


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Weijia Shang
%A Jose A. B. Fortes
%Z Purdue
%T On The Independent Partitioning of Algorithms With Uniform Data
Dependencies
%J PROC of the 1988 ICPP
%V II, Software
%I Penn State
%C University Park, Penn
%D AUG 1988
%P 26-33
%K Compilers, index set, dependence vectors,
greatest common divisor (GCD) method, minimum distance method, proof,


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Mary L. Bailey
%A David Socha
%A David Notkin
%Z U. Washington
%T Parallel Debugging Using Graphical Views
%J PROC of the 1988 ICPP
%V II, Software
%I Penn State
%C University Park, Penn
%D AUG 1988
%P 46-49
%K Software Tools, Voyeur, MIMD, shared memory,
icon, vector simulator views, X window system, micro-VAX,


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Pradip Bose
%Z IBM TJW
%T Heuristic, Rule-Based Program Transformations for Enhanced
Vectorization
%J PROC of the 1988 ICPP
%V II, Software
%I Penn State
%C University Park, Penn
%D AUG 1988
%P 63-66
%K Software Tools, IBM 3090 VF, Expert Advisor for VEctorization (EAVE),
RPTOOL, parallel Fortran (PF), loop interchange (restructuring),


%X /u1/rrh/bib/rrh.d/parallel.ref
%A Ron Cytron
%A Steve Karlovsky
%A Kevin P. McAuliffe
%Z IBM TJW
%T Automatic Management of Programmable Caches
%J PROC of the 1988 ICPP
%V II, Software
%I Penn State
%C University Park, Penn
%D AUG 1988
%P 229-238
%K Compilers, shared memory, PTRAN, partitioning, data dependence,
scheduling/direction vector, flush (post and invalidate), data flow,
optimization, eviction, line size, processor allocation,
%Z Extended abstract. This paper also uses different terminoogy
than some readers many have used.


%X /u1/rrh/bib/rrh.d/parallel.ref
%T Vector Optimization vs. Vectorization
%A M. Wolfe
%J PROC 1st INTL CONF on Supercomputing, LNCS 297
%C Athens, Greece
%D JUN 1987
%P 309-315
%X -peterd 8/25/89


%X /u1/rrh/bib/rrh.d/parallel.ref
%T Access Patterns: A Useful Concept in Vector Programming
%A Y. Jegou
%J PROC 1st INTL CONF on Supercomputing, LNCS 297
%C Athens, Greece
%D JUN 1987
%P 377-391
%X -peterd 8/30/89


%X /u1/rrh/bib/rrh.d/hardware.ref
%T Second Generation of Vector Supercomputers
%A E.W. Kozdrowicki
%A Douglas J. Theis
%J Computer
%V 13
%N 11
%D November 1980
%P 71-83


%X /u1/rrh/bib/rrh.d/hardware.ref
%T Special Tutorial: Vector Supercomputers
%A Douglas J. Theis
%J Computer
%V 7
%N 4
%D April 1974
%P 52-61
--


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.