treml@informatik.tu-muenchen.de (Thomas Treml) (01/10/91)
FINAL PROGRAM
Gesellschaft fuer Informatik
Informationstechnische Gesellschaft im VDE
in cooperation with IFIP
sponsored by ERCIM
The Second European
Distributed Memory Computing Conference
EEEEEE DDDDD M M CCCCC CCCCC 22222
EE DD DD MMM MMM CCCC CCCC 22 222
EE DD DD MM MM MM MM CC CC 22
EEEE DD DD MM MMM MM CC CC 2222
EE DD DD MM MM CC CC 222
EE DD DD MM MM CCCC CCCC 22
EEEEEE DDDDD MM MM CCCCC CCCCC 22222222
(EDMCC2)
22 - 24 April 1991
Institut fuer Informatik
D-8000 Muenchen 2 (FRG)
PRESENTATION
Todays applications with demand for very high performance,
fault tolerance and good price/performance ratio have triggered
the development of a large number of massively parallel archi-
tectures for commercial or research use. For such architectures,
totally distributed physical resources including memory are a
must in order to avoid the bottleneck of a globally shared
resource. On top of such physically distributed resources,
virtual global schemes (interconnection, memory, I/O) will be
offered in some future systems. A number of different inter-
connection topologies have been proposed including mesh, tree,
ring, hypercube, hierarchical organizations and multistage
networks.
Such distributed computers are now being used in European
industrial and public research and development centers. The
aim of this conference is to bring together users and devel-
opers to exchange their views and experiences in this field.
After a first workshop with the focus on hypercube architec-
tures organized by INRIA and IRISA in Rennes, France in 1989,
this second conference has changed its name slightly to
broaden the view to all distributed memory computers. This
second conference EDMCC2 is organized jointly by GI, Gesell-
schaft fuer Informatik, and ITG, Informationstechnische Gesell-
schaft im VDE, in cooperation with IFIP, International Federa-
tion for Information Processing, and SFB 342, Methods and
tools for the use of parallel architectures at the Technische
Universitaet Muenchen, FRG.
EDMCC2 Final Program
22 April 1991
10:00 h - 10:30 h Opening Remarks
10:30 h - 11:30 h Invited Lecture:
Justin Rattner (ISC)
The New Age of Supercomputing
11:30 h - 12:30 h Invited Lecture:
David May (Inmos)
The Next Generation Transputers and Beyond
12:30 h - 13:30 h Lunch
13:30 h - 15:30 h Parallel Sessions
Session 1.1: Virtual Shared Memory
W.K. Giloi, C. Hastedt, F. Schoen, W.
Schroeder-Preikschat, (D)
A Distributed Implementation of Shared Virtual Memory
with Strong and Weak Coherence
L. Borrmann, P. Istavrinos (D)
Store coherency in a parallel distributed-memory
machine
C. Hammer, T. Henties (D)
Using a Weak Coherency Model for a Parallel Lisp
P.K. Sinha, H. Ashihara, K. Shimizu, M. Maekawa (Japan)
Flexible User-Definable Memory Coherence Scheme in
Distributed Shared Memory of GALAXY
Session 1.2: Tools (1)
M.H. Gerndt (A)
Parallelization of Multigrid Programs in SUPERB
F. Lamour, P. Chretienne (F)
Mapping graphs onto a partially reconfigurable
architecture
B. Mohr (D)
SIMPLE: a performance evaluation tool environment for
parallel and distributed systems
A. Fernandez, J.M. Llaberia, J.J. Navarro, M.
Valero Garcia (E)
Interleaving Partitions of Systolic Algorithms for
Programming Distributed Memory Multiprocessors
15:30 h - 16:00 h Coffee Break
16:00 h - 18:00 h Parallel Sessions
Session 2.1: Interconnection Problems (1)
H. Scheidig, M.F. Schneider, R. Spurk (D)
Efficient and Scalable Logical Busses for Message
Passing Interconnection Networks
H. Hofestaedt, A. Klein, E. Reyzl (D)
Performance Benefits from Locally Adaptive Interval
Routing in Dynamically Switched Interconnection
Networks
Xie Xiaohong, Ge Youmei (UK)
An Optimal Structure that Accomodates Both a Ring
and a Binary Tree
H. Corporaal, J.G.E. Olk (NL)
A Scalable Communication Processor Design supporting
Systolic Communication
Session 2.2: Applications (1)
C. Aykanat, T. Kurc, F. Ercal (Turkey)
Parallelization of Lee's Routing Algorithm on a Hyper-
cube Multicomputer
T. Bemmerl, U. Graf, R. Knoedlseder (D)
Experiences in parallelizing an existing CFD algorithm
A.R. Robertson, R.N. Ibbett (UK)
Simulation of the MC88000 Microprocessor System on a
Transputer Network
L. Bacchelli Montefusco, C. Guerrini (I)
A Domain Decomposition Method for Scattered Data
Approximation on a Distributed Memory Multiprocessor
23 April 1991
9:00 h - 10:30 h Parallel Sessions
Session 3.1: Systems Software
M. Raynal, A. Couvert, R. Pedrono (F)
Implementation and Evaluation of Distributed synchro-
nization on a distributed memory parallel machine
E. Leu, A. Schiper, A. Zramdini (CH)
Efficient Execution Replay Technique for Distributed
Memory Architectures
H. Corporaal (NL)
Distributed Heapmanagement using reference weights
Session 3.2: Tools (2)
G.J.W. van Dijk, A.J. van der Wal (NL)
Partial Ordering of Synchronization Events for Distri-
buted Debugging in Tightly-coupled Multiprocessor
Systems
S. Chaumette, M.C. Counilh (F)
A Development for Distributed Systems
C.-W. Oehlrich, P. Metzger, A. Quick (D)
Monitor-Supported Analysis of a Communication System
for Transputer-networks
10:30 h - 11:00 h Break
11:00 h - 12:30 h Panel Discussion
12:30 h - 13:30 h Lunch
13:30 h - 15:30 h Parallel Sessions
Session 4.1: Programming Languages and Algorithms
J.-L. Pazat (F)
Code Generation for Data Parallel Programs on DMPCs
J.-L. Giavitto, C. Germain, J. Fowler (F)
OAL: an Implementation of an Actor Language on a
Massively Parallel Message-Passing Architecture
A. Cheese (D)
Implementing Committed-Choice Logic Programming
Languages on Distributed Memory Computers
S.A. Delgado-Rannauro, M. Dorochevsky, K. Schuerman,
A. Veron, J. Xu (D)
A Shared Environment Parallel Logic Programming
System on Distributed Memory Architectures
D. Gomm, M. Heckner, K.-J. Lange, G. Riedle (D)
On the Design of Parallel Programs for Machines with
Distributed Memory
Session 4.2: Alternative Execution Models
W. Schreiner (A)
ADAM - An Abstract Dataflow Machine and its
Transputer Implementation
P. Clauss (F)
Mapping Systolic Algorithms on Distributed Memory
Computers
C. Percebois, N. Signes, P. Agnoletto (F)
A Compiler for a Distributed Inference Model
P. Sapaty (USSR)
Logic Flow in Active Data as a New Model for Distri-
buted Knowledge Processing
15:30 h - 16:00 h Coffee Break
16:00 h - 19:00 h Poster Session
20:00 h Official Dinner
24 April 1991
9:00 h - 10:30 h Parallel Sessions
Session 5.1: Applications (2)
S. Noelting (D)
Nonlinear Adaptive Finite Element Systems on
Distributed Memory Computers
O. Zone, R. Keunings, D. Roose (B)
Parallel Algorithms for the Direct Solution of Finite
Element Equations on a Distributed Memory Computer
A.H. Nelson (UK)
Simulations of Galaxy Interactions on a massively
Parallel Computer
Session 5.2: Tools (3)
T. Bemmerl, A. Bode (D)
An Integrated Environment for Programming Distributed
Memory Multiprocessors
A. Touzene, B. Plateau (F)
Optimal Multinode Broadcast on a Mesh Connected
Graph with Reduced Bufferization
J. De Keyser, D. Roose (B)
Adaptive Irregular Multiple Grids on a Distributed
Memory Multiprocessor
10:30 h - 11:00 h Break
11:00 h - 12:30 h Parallel Sessions
Session 6.1: Operating Systems and Related Topics
H. Schmidt (D)
Making PEACE a Dynamic Alterable System
S. Nicolle, P. Legrand, J.-P. Derutin (F)
Managing the recursive Generation of Tasks in a
Transputer Network
B. Herrmann, L. Philippe (F)
Multicomputers UNIX based on CHORUS
Session 6.2: Systems (1)
U. Bruening, W.-K. Giloi, W. Schroeder-Preikschat (D)
The Architecture of the European MIMD Supercomputer
GENESIS
P. Fraigniaud (F)
Fault-tolerant gossiping on hypercube multicomputers
M. Annaratone, G. zur Bonsen, M. Fillo, M. Halbherr,
R. Ruehl, P. Steiner, M. Viredaz (CH)
Architecture, Implementation, and System Software of
K2
12:30 h - 13:30 h Lunch
13:30 h - 14:30 h Parallel Sessions
Session 7.1: Interconnection Problems (2)
V.S. Getov, C. Jesshope (UK)
Simulation Facility of Distributed Memory System with
"Mad Postman" Communication Network
J. Duato (E)
On the Design of Deadlock-Free Adaptive Routing
Algorithms for Multicomputers: Theoretical Aspects
Session 7.2: Systems (2)
G. Watzlawik, E.H. Robinson (D)
European Declarative System (EDS): Architecture and
Interprocess Communication
A. Bauch, R. Braam, E. Maehle (D)
DAMP - A Dynamic Reconfigurable Multiprocessor
System with a Distributed Switching Network
14:30 h - 15:00 h Break
15:00 h - 16:00 h Session 8.1: Tools (4)
T. Ungerer, L. Bic (D, USA)
An Object-Oriented Interface for Parallel Programming
of Loosely-Coupled Multiprocessor Systems
J. Torres, E. Ayguade, J. Labarta, J.M. Llaberia, M.
Valero (E)
On Automatic Loop Data-Mapping for Distributed-
Memory Multiprocessors
17:00 h Closing Remarks
Registration
The registration desk will be open Monday through Wednesday 8:00
am to 5:00 pm.
Conference Secretariate
Secretariate EDMCC2
Prof. Dr. Arndt Bode
Institut fuer Informatik
Technische Universitaet Muenchen
Postfach 20 24 20, Arcisstr. 21
D-8000 Muenchen 2
Germany
Tel.: +49-89-2105-8240
FAX: +49-89-2800529
email: bode@lan.informatik.tu-muenchen.de
Hotel and Accomodation
Hotel reservations are exclusively handled by the Munich Tourist Office.
Deadline for reservations is 15 March 1991.
Landeshauptstadt Muenchen
Fremdenverkehrsamt
Postfach
D-8000 Muenchen 1
Germany
Lunch
The TU Munich Mensa offers meals at modest prices. Various restaurants are
within walking distance.
Exhibition and Posters
Additionally to the regular sessions of the conference there will be an
exhibition and presentation of vendor products and research in the field of
multiprocessor systems (until now the following vendors have decided to
participate: ALLIANT, GENIAS, INTEL SUPERCOMPUTER DIVISION, MEIKO,
NCUBE etc.). The exhibition will be freely accessible to the public, it is
taking place within the main building, just outside the auditorium and will be
open daily from 9:00 am to 6:00 pm.
Projects not included into the regular sessions will be presented as posters.
All posters will be exhibited throughout the conference, a special poster
session will give the authors of posters the opportunity to discuss their work
with interested participants.
Registration procedure
To register, fill out the following form:
-------------------------------------------------------------------------------
__
Name:___________________________________ |__| Member GI, AFCET,
AICA, NGI, ITG, ACM
First Name:_____________________________ Member No: __________
(DM 290,-- before 31
Title:__________________________________ January 1991, DM 350,--
after)
__
|__| Nonmember
Institution:____________________________ (DM 390,-- before 31
January 1991, DM 450,--
after)
__
Address:________________________________ |__| Student
(DM 50,-- before 31
________________________________________ January 1991, DM 100,--
after)
________________________________________ The conference fee
__
________________________________________ |__| has been transfered to the
following account:
________________________________________ Prof. Dr. Arndt Bode,
EDMCC2
________________________________________ Dresdner Bank Muenchen
Account No.: 5082 444 01
Bank code: 700 800 00
__
Telephone:_______________________________ |__| is included as a check
enclosed to this letter
Signature:_______________________________
--------------------------------------------------------------------------
Send it to:
Prof. Dr. Arndt Bode
EDMCC2
Institut fuer Informatik
Technische Universitaet Muenchen
Postfach 20 24 20
D-8000 Muenchen 2
Germany
Conference Chairman
Prof. Dr. Arndt Bode
Program committee members
C. Addison
F. Andre
K. Antreich
T. Bemmerl
A. Bode
W. Brauer
R. Chamberlain
M. Cosnard
E. Dagless
W. Giloi
W. Haendler
R. Hempel
F. Hofmann
C. Jesshope
B. Kagstroem
P. Leca
P. Mueller-Stoy
J. Nehmer
T. Priol
P. Quinton
K. Reinartz
G.L. Reijns
D. Roose
T. Schwederski
M. Valero
M. Vanneschi
J.P. Verjus
K. Waldschmidt
C. Whitby-Strevens