PPOPP93 program announcement
Marina Chen
This announcement contains, in addition to the Advance Program
of PPoPP93, information about Federated Computing Research
Conference '93 of which PPoPP is a participating conference.
See enclosed instructions for access the on line FCRC program
and registration information at the end of this announcement.
==================================================================
Advance Program
Fourth ACM SIGPLAN Symposium on
Principles and Practices of Parallel Programming
Sheraton Harbor Island Hotel
San Diego, California, May 20-22, 1993
===================================================================
===================================================================
PRECONFERENCE TUTORIAL SESSIONS on May 19, 1993
===================================================================
TUTORIAL A.
1. Lecturer: Geoffrey C. Fox
2. Affiliation: Professor of Computer Science and Physics at Syracuse
University, and Director of the Northeast Parallel Architectures Center.
3. Date: May 19
4. Time (start-end): 9:20 -- 10:35 Coffee Break 11:05 -- 12:20
5. Room: Bel Aire North (Sheraton Grand Hotel)
6. Title: "High Performance Computing Enabling Technologies for the
Real World"
7. Description: We present and discuss the results of a survey of the
relevance of parallel computing in New York State Industry. We divide
applications into 18 fields varying broadly from large scale simulation
to information processing and including the support of an interactive
virtual world for the state's children. We then examine software and
algorithmic technologies which appear to be needed to support the
different applications. Software systems include databases, languages
(high performance Fortran and C++) and the integration software at
various levels, including tele-virtuality operating environments.
Algorithms include matrix algebra, irregular mesh generation and large
scale scheduling methods. We discuss which application areas appear
to need which technologies.
TUTORIAL B.
1. Lecturer: David A. Padua
2. Affiliation: Associate Director of the Center for Supercomputing
Research and Development, and Associate Professor in the Department
of Computer Science at the University of Illinois at Urbana-Champaign.
3. Date: May 19
4. Time (start-end): 9:20 -- 10:35 Coffee Break 11:05 -- 12:20
5. Room:
6. Title: Compiler Techniques for Parallel and Vector Computers
7. Description: Cabernet/Chardonnay
The tutorial will present an overview of the issues that
arise in the development of compilers for parallel and vector
computers. Both shared- and distributed-memory machines will be
discussed. The source languages considered include both sequen-
tial languages such as Fortran 77, extended with compiler asser-
tions, and parallel dialects such as PCF Fortran and HPF Fortran.
The transformation and analysis techniques to be covered
include:
(1) Dependence analysis of sequential code and conflict analysis
of parallel code.
(2) Transformations to detect and exploit implicit parallelism
and to enhance explicit parallelism. These include loop
transformation techniques such as partitioning and distribu-
tion and other strategies such as induction variable recog-
nition and variable renaming.
(3) Transformations to decrease communication and control over-
head, to enhance the locality of parallel loops, and algo-
rithms to generate efficient code for synchronization, com-
munication, and prefetching.
(4) Compile-time data manipulation techniques such as privatiza-
tion, alignment and data partitioning across memory modules.
(5) Run-time techniques for scheduling and parallelism detection
and their relationship to language constructs and transla-
tion strategies.
Experimental measurements of the effectiveness of several
of the translation and analysis techniques will be presented
throughout the tutorial.
TUTORIAL C.
1. Lecturer: Guy L. Steele Jr.
2. Affiliation: Thinking Machines Corporation
3. Date: May 19
4. Time (start-end): 2:15 -- 3:30 Coffee Break 4:00 -- 5:15
5. Room: Bel Aire North (Sheraton Grand Hotel)
6. Title: High Performance Fortran
7. Description:
The goal of High Performance Fortran (HPF) is to extend Fortran
90 to provide additional support for data parallel programming
(defined as a style of programming with a single conceptual
thread of control, a global name space, and loosely synchronous
parallel computation) to facilitate top performance on MIMD and
SIMD computers with non-uniform memory access costs, while also
promoting, or at least not impeding, performance on other
machines. The idea is to promote portability of Fortran
programs over a larger class of computers: distributed-memory
MPP (massively parallel processor), shared-memory MPP, vector
computers, multihead vector computers, shared-memory
multicomputers, mainframes, and workstations.
In this tutorial we will cover all aspects of the definition of
High Performance Fortran, including directives for controlling
data alignment and distribution; the FORALL statement; new
intrinsic and library functions; extrinsic and local procedures;
the treatment of storage association, COMMON, and EQUIVALENCE;
and the relationship of High Performance Fortran to Fortran 90.
TUTORIAL D.
1. Lecturers: Tom LeBlanc and Mark Crovella
2. Affiliation: Both of the University of Rochester, where LeBlanc is
an Associate Professor of Computer Science and Crovella is a
Ph.D. candidate in Computer Science.
3. Date: May 19
4. Time (start-end): 2:15 -- 3:30 Coffee Break 4:00 -- 5:15
5. Room: Cabernet/Chardonnay
6. Title: "Runtime Tools for Parallel Programming"
7. Description:
This tutorial will describe tools and techniques for the dynamic
(runtime) analysis of parallel programs. The discussion will first
present the foundational concepts related to debugging and performance
tuning of parallel programs; then the talk will explore those concepts
using a survey of available tools. The concepts to be discussed
include: parallel execution as a partial order; one time vs. cyclic
execution analysis; intrusive vs. non-intrusive monitoring; tradeoffs
between monitoring overhead and precision; performance metrics for
parallel programs; sources of inefficiency in parallel execution; and
techniques for displaying debugging and performance information. The
tool survey will examine how these concepts affect the design and
utility of representative tools. We will examine tools from four
categories: profiling tools (e.g, Pixie, Quartz, Forge); tools for event
monitoring and display (e.g., Gist, Pablo, Instant Replay); tools for
analyzing memory hierarchy usage (e.g. MTool); and tool systems for
debugging and performance analysis (e.g, IPS, PPuTTs).
===================================================================
CONFERENCE SESSIONS
===================================================================
1. Title Session 1
2. Date May 20, 1993
3. Time 9:20 - 10:50
4. Room Palomar
5. Chair Aki Yonezawa
6. Affiliation University of Tokyo
Papers:
1. Time 9:20 - 9:50
2. Paper title LogP: Towards a Realistic Model of Parallel Computation
3. Author D. Culler, R. Karp, D. Patterson, A. Sahay, K.E. Schauser,
E. Santos, R. Subramonian, and T. von Eicken
4. Affiliation University of California at Berkeley
1. Time 9:50 - 10:20
2. Paper title Programming Task and Data Parallelism on a Multicomputer
3. Author J. Subhlok, J.M. Stichnoth, D.R. O'Hallaron, and T. Gross
4. Affiliation Carnegie Mellon University
1. Time 10:20 - 10:50
2. Paper title ActorSpaces: An Open Distributed Programming Paradigm
3. Author G. Agha and C. Callsen
4. Affiliation University of Illinois at Urbana-Champaign
---------------------------------------------------------------------------------
1. Title Session 2
2. Date May 20, 1993
3. Time 11:20 - 12:20
4. Room: Palomar
5. Chair Dan Reed
6. Affiliation University of Illinois at Urbana-Champaign
Papers:
1. Time 11:20 - 11:50
2. Paper title Experiences Using the ParaScope Editor: an Interactive
Parallel Programming Tool
3. Author M.W. Hall, T.J. Harvey, K. Kennedy, N. McIntosh, K.S.
McKinley, J.D. Oldham, M.H. Paleczny, and G. Roth
4. Affiliation Stanford University, Rice University, Rice University,
Rice University, ENSMP (CRI), Stanford University,
Rice University, and Rice University
1. Time 11:50 - 12:20
2. Paper title Perturbation Analysis of High Level Instrumentation for
SPMD Programs
3. Author S.R. Sarukkai and A. Malony
4. Affiliation Indiana University and University of Oregon
---------------------------------------------------------------------------------
1. Title Session 3
2. Date May 20, 1993
3. Time 2:00 - 4:00
4. Room: Palomar
5. Chair Monica Lam
6. Affiliation Stanford University
Papers:
1. Time 2:00 - 2:30
2. Paper title Integrating Message-Passing and Shared-Memory: Early
Experience
3. Author D. Kranz, K. Johnson, A. Agarwal, J. Kubiatowicz, and
B.-H. Lim
4. Affiliation MIT
1. Time 2:30 - 3:00
2. Paper title Using Scheduler Information to Achieve Optimal Barrier
Synchronization Performance
3. Author L. Kontothanassis and R. Wisniewski
4. Affiliation University of Rochester
1. Time 3:00 - 3:30
2. Paper title A Concurrent Copying Garbage Collector for Languages that
Distinguish (Im)mutable Data
3. Author L. Huelsbergen and J.R. Larus
4. Affiliation University of Wisconsin - Madison
1. Time 3:30 - 4:00
2. Paper title Improving the Performance of Runtime Parallelization
3. Author S.-T. Leung and J. Zahorjan
4. Affiliation University of Washington
--------------------------------------------------------------------------------
1. Title Panel session
2. Date May 20, 1993
3. Time 4:30 - 6:00
4. Room Palomar
5. Chair (to be determined)
6. Affiliation
-------------------------------------------------------------------------------
1. Title Session 4
2. Date May 21, 1993
3. Time 9:20 - 10:50
4. Room Palomar
5. Chair Guy Steele
6. Affiliation Thinking Machines Corporation
Papers:
1. Time 9:20 - 9:50
2. Paper title High Performance Fortran Without Templates: A New Model
for Distribution and Alignment
3. Author B.M. Chapman, P. Mehrotra, and H.P. Zima
4. Affiliation University of Vienna, ICASE, and University of Vienna
1. Time 9:50 - 10:20
2. Paper title Implementation of a Portable Nested Data-Parallel Language
3. Author G.E. Blelloch, S. Chatterjee, J. Hardwick, J. Sipelstein,
and M. Zagha
4. Affiliation Carnegie Mellon University, RIACS, Carnegie Mellon University,
Carnegie Mellon University, and Carnegie Mellon University
1. Time 10:20 - 10:50
2. Paper title An Equational Language for Data-Parallelism
3. Author P. Rao and C. Walinsky
4. Affiliation Dartmouth College
-------------------------------------------------------------------------------
1. Title Session 5
2. Date May 21, 1993
3. Time 11:20 - 12:20
4. Room Palomar
5. Chair Hans Zima
6. Affiliation University of Vienna
Papers:
1. Time 11:20 - 11:50
2. Paper title Transforming High-Level Data-Parallel Programs into Vector
Operations
3. Author J.F. Prins and D.W. Palmer
4. Affiliation University of North Carolina
1. Time 11:50 - 12:20
2. Paper title Non-concurrency Analysis
3. Author S.P. Masticola and B.G. Ryder
4. Affiliation Rutgers University
-------------------------------------------------------------------------------
1. Title Session 6
2. Date May 21, 1993
3. Time 2:00 - 3:30
4. Room Palomar
5. Chair Michael Wolfe
6. Affiliation Oregon Graduate Institute
Papers:
1. Time 2:00 - 2:30
2. Paper title Explicit Data Placement (XDP): A Methodology for Explicit
Compile-Time Representation and Compile-Time Representation
and Optimization of Data Movement
3. Author V. Bala, J. Ferrante, and L. Carter
4. Affiliation IBM T.J. Watson Research Center
1. Time 2:30 - 3:00
2. Paper title Generating Local Address and Communication Sets for
Data-Parallel Programs
3. Author S. Chatterjee, J.R. Gilbert, F.J.E. Long, R. Schreiber,
and S.-H. Teng
4. Affiliation RIACS, Xerox PARC, University of California at Santa Cruz,
RIACS, and MIT
1. Time 3:00 - 3:30
2. Paper title Data Flow Equations for Explicitly Parallel Programs
3. Author D. Grunwald and H. Srinivasan
4. Affiliation University of Colorado at Boulder
--- ---------------------------------------------------------------------------
1. Title Session 7
2. Date May 21, 1993
3. Time 4:00 - 5:30
4. Room Palomar
5. Chair Olaf Lubeck
6. Affiliation Los Alamos National Laboratory
Papers:
1. Time 4:00 - 4:30
2. Paper title Implementing an Irregular Application on a Distributed
Memory Multiprocessor
3. Author S. Chakrabarti and K. Yelick
4. Affiliation University of California at Berkeley
1. Time 4:30 - 5:00
2. Paper title Parallelizing Algorithms for Symbolic Computation using
||MAPLE||
3. Author K. Siegl
4. Affiliation Johannes Kepler University
1. Time 5:00 - 5:30
2. Paper title Experience with Fine-Grain Synchronization in MIMD
Machines for Preconditioned Conjugate Gradient
3. Author D. Yeung and A. Agarwal
4. Affiliation MIT
-------------------------------------------------------------------------------
1. Title PPoPP business meeting
2. Date May 21, 1993
3. Time 5:30 - 6:00
4. Room Palomar
5. Chair
6. Affiliation
-------------------------------------------------------------------------------
1. Title Session 8
2. Date May 22, 1993
3. Time 9:20 - 10:50
4. Room Palomar
5. Chair Robert Halstead
6. Affiliation DEC Cambridge Research Lab
Papers:
1. Time 9:20 - 9:50
2. Paper title Procs and Locks: A Portable Multiprocessing Platform for
Standard ML of New Jersey
3. Author J.G. Morrisett and A. Tolmach
4. Affiliation Carnegie Mellon University and Portland State University
1. Time 9:50 - 10:20
2. Paper title Leapfrogging: A Portable Technique for Implementing
Efficient Futures
3. Author D.B. Wagner and B.G. Calder
4. Affiliation University of Colorado at Boulder
1. Time 10:20 - 10:50
2. Paper title An Efficient Implementation Scheme of Concurrent
Object-Oriented Languages on Stock Multicomputers
3. Author K. Taura, S. Matsuoka, and A. Yonezawa
4. Affiliation University of Tokyo
-------------------------------------------------------------------------------
1. Title Session 9
2. Date May 22, 1993
3. Time 11:05 - 12:35
4. Room Palomar
5. Chair David Callahan
6. Affiliation Tera Computer Company
Papers:
1. Time 11:05 - 11:35
2. Paper title The Shared Regions Approach to Software Cache Coherence on
Multiprocessors
3. Author H.S. Sandhu, B. Gamsa, and S. Zhou
4. Affiliation University of Toronto
1. Time 11:35 - 12:05
2. Paper title Computation Migration: Enhancing Locality for
Distributed-Memory Parallel Systems
3. Author W. Hsieh, P. Wang, and W.E. Weihl
4. Affiliation MIT
1. Time 12:05 - 12:35
2. Paper title Data Locality and Load Balancing in COOL
3. Author R. Chandra, A. Gupta, and J.L. Hennessy
4. Affiliation Stanford University
===================================================================
Federated Computing Research Conference '93
CRA, with support from the National Science Foundation, has taken the lead in
organizing the first Federated Computing Research Conference, which will be
held May 15-21, 1993, at the Sheraton Harbor Island Hotel in San Diego. By
providing a common time and meeting place for several established research
meetings, FCRC '93 is creating a major research conference that will not lose
the scientific benefits of the computing research field's smaller, more
specialized meetings.
FCRC '93 is retaining the intellectual benefits and research identities of the
smaller constituent meetings, while providing greater visibility for the field.
FCRC '93 also is providing the opportunity for researchers to meet with their
peers in other specialties. Because of the unified nature of the conference,
more people in the computing research community will be able to learn about
important findings in the specialized subfields.
Each participating conference will be independently administered, and each
organizing group will be responsible for their meeting's structure, content and
proceedings. All attendees will register for at least one participating
conference and will be able to buy the proceedings from the other meetings.
During their "home" conference, to the extent facilities will allow, attendees
will be free to sit in on other meetings.
Each morning will start with a plenary talk on a topic of broad appeal to the
community. The conference features two plenary evening social events.
In planning FCRC '93, CRA received support and assistance from the other
sponsoring organizations. The Association for Computing Machinery has been
particularly active in providing support and planning expertise. For more
information on FCRC '93, contact CRA at 202-234-2111.
Participating Research Meetings
Confirmed participants in FCRC '93 are as follows (Other smaller, ad-hoc
workshops also are being added to the program.):
* 25th Annual ACM Symposium on the Theory of Computing (STOC)
Sponsor: ACM Special Interest Group on Algorithms and
Computation Theory (SIGACT)
Contact: David S. Johnson, AT&T Bell Labs, d...@research.atl.com
* Ninth Annual ACM Symposium on Computational Geometry
Sponsors: ACM SIGACT and ACM Special Interest Group on Graphics
(SIGGRAPH)
Contact: Chee Yap, Courant Institute, y...@yap.cs.nyu.edu
* Fourth ACM SIGPLAN Symposium on Principles and Practices of Parallel
Programming (PPoPP)
Sponsor: ACM Special Interest Group on Programming Languages
(SIGPLAN)
Contact: Marina Chen, Yale University, chen-...@cs.yale.edu
* Eighth Annual Conference on Structure in Complexity Theory
Sponsor: IEEE Technical Committee on Mathematical Foundations
of Computing
Contact: Steve Mahaney, University of Arizona,
s...@cs.arizona.edu
* Workshop on Parallel Algorithms (WOPA '93)
Sponsor: University of Maryland Institute for Advanced Computer
Studies (UMIACS) and the Defense Advanced Research Projects Agency
Contact: Uzi Vishkin, University of Maryland,
vis...@umiacs.umd.edu
* 20th Annual International Symposium on Computer Architecture (ISCA '93)
Sponsors: ACM Special Interest Group on Computer Architecture,
IEEE Computer Society and the IEEE-CS Technical Committee on Computer
Architecture (TCCA)
Contact: Lubomir Bic, University of California at Irvine,
b...@cj2.ics.uci.edu
* Seventh Workshop on Parallel and Distributed Simulation (PADS)
Sponsors: ACM Special Interest Group on Simulation (SIGSIM),
IEEE Computer Society, IEEE-CS Technical Committee on Simulation (TCSIM) and
the Society for Computer Simulation (SCS)
Contact: Richard Fujimoto, Georgia Institute of Technology
fuji...@cc.gatech.edu
* ACM/ONR Workshop on Parallel and Distributed Debugging
Sponsors: Office of Naval Research, ACM SIGPLAN and the ACM
Special Interest Group on Operating Systems
Contacts: Bart Miller, University of Wisconsin, ba...@cs.wisc.edu
and Joan Francioni, University of Southwestern Louisiana, j...@cacs.usl.edu
* CRA Workshop on Academic Careers for Women
Sponsor: CRA's Committee on the Status of Women
Contact: Cynthia Brown, Northeastern University,
br...@corwin.ccs.northeastern.edu
*************************************
CONFERENCE SCHEDULE
5-14 5-15 5-16 5-17 5-18 5-19 5-20 5-21 5-22
FRI SAT SUN MON TUE WED THU FRI SAT
----- ----- ----- ----- ----- ----- ----- ----- -----
ISCA WS WS T |<----------->|
STOC R |<------------->|
PADS T/R |<------------->|
WOPA |<-RP---->|
PPOPP T/R |<----------->|
CS Str R |<------------------->|
Cmp Geom R |<------------->|
WPDD |<------->|
CRA WS |<->|
*************************************
R=reception; T=tutorial; WS=workshop
ISCA=20th Annual International Symposium on Computer Architecture
STOC=25th Annual ACM Symposium on the Theory of Computing
PADS=Seventh Workshop on Parallel and Distributed Computing
WOPA=Workshop on Parallel Algorithms
PPOPP=Fourth ACM SIGPLAN Symposium on Principles and Practices of Programming
CS Str=Eighth Annual Conference on Structure in Complexity Theory
Cmp Geom=Ninth Annual ACM Symposium on Computational Geometry
WPDD=ACM/ONR Workshop on Parallel and Distributed Debugging
CRA WS=CRA Workshop on Academic Careers on Academic Careers for Women
*********************************************************************
Postscript files of the updated program and registration form are available
by anonymous FTP (use "anonymous" as login name and your full email address
as password) from
cs.purdue.edu:/pub/fcrc
The files are
program.ps -- the whole program, including the registration form
registrationForm.ps -- the registration form only
and the compressed versions of each. Note: The compressed files (*.Z) should
be transfered in "binary" mode.
*********************************************************************