Conference Program


Tuesday, July 8

Opening: 9:00 -- 9:30, Room A
Invited Talk: 9:30 - 10:30, Room A
Chair: Steve Wallach

The Characteristics of the Uniprocessor in Future Supercomputers
Yale N. Patt, Univ. of Michigan, MI, USA
Coffee Break: 10:30 - 11:00
Session 1a: 11:00 - 12:00, Room A
Instruction-Level Parallelism and Wide-Busses
Chair: Trevor Mudge

Session 1b: 11:00 - 12:00, Room B
Collective I/O
Chair: Peter Brezany

Scalable Instruction-Level Parallelism
Through Tree Instructions
Jaime H. Moreno and Mayan Moudgill
IBM T. J. Watson Research Center, Yorktown Heights, NY, USA
Implementation of Collective I/O in the Intel Paragon Parallel File System: Initial Experiences
Rajesh Bordawekar
California Institute of Technology, Pasadena, CA, USA
Increasing Memory Bandwidth With Wide Buses: Compiler, Hardware and Performance Trade-Offs
David Lopez, Mateo Valero, Josep Llosa, and Eduard Ayguade
Univ. Politecnica de Catalunya, Barcelona, Spain
Optimizing Collective I/O Performance on Parallel Computers: A Multisystem Study
Ying Chen*, Ian Foster***, Jarek Nieplocha**, and Marianne Winslett*
*Univ. of Illinois, Urbana, IL, USA; **Pacific Northwest National Lab., Richland, WA, USA; ***Argonne National Lab., Argonne, IL, USA
Invited Talk: 12:00 -- 12:30, Room A
Chair: Hans Zima

HPCN in the ESPRIT Programme
Khalil Rouhana, European Commission, Brussels
Lunch Break: 12:30 - 14:00
Session 2a: 14:00 - 15:30, Room A
Applications / Scheduling
Chair: Christoph Ueberhuber

Session 2b: 14:00 - 15:30, Room B
Caches
Chair: Josep Torrellas

Performance Improvement Through Overhead Analysis: A Case Study in Molecular Dynamics
Graham D. Riley, J. Mark Bull, and John R. Gurd
University of Manchester, UK
Data Caches for Superscalar Processors
Toni Juan, Juan J. Navarro, Olivier Temam*
Univ. Politecnica de Catalunya, Barcelona, Spain; *Versailles Univ., France
Generalized Cannon's Algorithm for Parallel Matrix Multiplication
Hyuk-Jae Lee, James P. Robertson, and Jose A.B. Fortes
Louisiana Tech. Univ., Ruston, LA, USA
Improving Data Cache Performance by Pre-Executing Instructions Under a Cache Miss
James Dundas and Trevor Mudge
Univ. of Michigan, Ann Arbor, MI, USA
Multiprocessor Scheduling with Client Resources to Improve the Response Time of WWW Applications
Daniel Andresen and Tao Yang
University of California, Santa Barbara,
CA, USA
Eliminating Cache Conflict Misses Through XOR-Based Placement Functions
Antonio Gonzalez, Mateo Valero, Nigel Topham* and Joan M. Parcerisa
Univ. Politecnica de Catalunya, Barcelona,
Spain; *Univ. of Edinburgh, UK
Coffee Break: 15:30 - 16:00
Panel Session: 16:00 - 18:00, Room A
Debate: On The Relative Merits Of Parallel Programming Using a Single-Address-Space Model Versus Message-Passing Between Multiple Address Spaces
Chair: John Gurd, Univ. of Manchester, UK
Reception - City of Vienna (City Hall): 20:00


Wednesday, July 9

Invited Talk: 8:30 - 9:30, Room A
Chair: Ralf Hofestädt

Computers in the Drug Discovery Cycle
Hwa Lim, Pangea Systems Inc., Oakland, CA, USA
Session 3: 9:30 - 10:30, Room A
Parallel Architectures
Chair: Mateo Valero

pSNOW: A Tool to Evaluate Architectural Issues for NOW Environments
Mangesh Kasbekar, Shailabh Nagar, Anand Sivasubramaniam
Pennsylvania State Univ., PA, USA
CP-PACS: A Massively Parallel Processor For Large Scale Scientific Calculations
T. Boku, K. Itakura, H. Nakamura and K. Nakazawa
Univ. of Tsukuba, Japan
Coffee Break: 10:30 - 11:00
Session 4a: 11:00 - 12:00, Room A
Object-Oriented Programming
Chair: Piyush Mehrotra

Session 4b: 11:00 - 12:00, Room B
Routing
Chair: Alok Choudhary

A Methodology for Specifying Data Distribution Using Only Standard Object-Oriented Features
Naohito Sato, Satoshi Matsuoka, Jean-Marc Jezequel, Akinori Yonezawa
Univ. of Tokyo, Japan
Impact of Selection Functions on Routing Algorithm Performance in Multicomputer Networks
Wu-chang Feng, Kang G. Shin
Univ. of Michigan, Ann Arbor, MI, USA
HPC++: Experiments with the Parallel Standard Template Library
Elizabeth Johnson, Dennis Gannon, and Peter Beckman*
Indiana Univ., Bloomington, IN, USA;
*Los Alamos National Lab., NM, USA
Performance Benefits of Virtual Channels And Adaptive Routing: An Application-Driven Study
Aniruddba S. Vaidya, Anand Sivasubramaniam, and Chita R. Das
Pennsylvania State Univ., PA, USA
Lunch Break: 12:00 - 14:00
Session 5a: 14:00 - 15:30, Room A
Synchronization
Chair: Michael Gerndt

Session 5b: 14:00 - 15:30, Room B
Performance
Chair: Thomas Fahringer

A Graph Based Approach to Barrier Synchronisation Minimisation
E.A. Stöhr and Michael F. P. O'Boyle
Univ. of Manchester, UK
Performance Evaluation of Message-Driven Parallel VLSI CAD Applications on General Purpose Multiprocessors
John G. Holm, John A. Chandy*, Steven Parkes*, Sumit Roy, Venkatram Krishnaswamy, Gagan Hasteer, and Prithviraj Banerjee**
University of Illinois, Urbana, IL, USA; *Sierra Vista Research, Los Gatos, CA, USA; **Northwestern University, Evanston, IL, USA
The Importance of Synchronization Structure in Parallel Program Optimization
Arjan J.C. van Gemund
Delft Univ. of Technology, The Netherlands
Incorporating Application Dependent Information In An Automatic Code Generating Environment
R. van Engelen, I. Heitlager, L. Wolters, and G. Cats*
Leiden University, The Netherlands; *Royal Netherlands Meteorological Institute, Ae de Bilt, The Netherlands

Sparse Code Generation for Imperfectly Nested Loops With Dependencies
Vladimir Kotlyar and Keshav Pingali
Cornell University, Ithaca, NY, USA
Coffee Break: 15:30 - 16:00
Panel Session: 16:00 - 18:00, Room A
Debate: On The Implementation Aspects of a Single-Address-Space Model Versus a Message-Passing Model As Seen From Independent Software Providers
Chair: Frank Baetke, Hewlett Packard, Munich, Germany
Conference Dinner (Palais Pallavicini): 20:00


Thursday, July 10

Invited Talk: 8:30 - 9:30, Room A
Chair: Harry Wijshoff

Problem Solving Environments and Netsolve: A Network Server for Solving Computational Science Problems
Jack Dongarra, Univ. of Tennessee, Knoxville, TN, USA
Session 6a: 9:30 - 10:30, Room A
Prefetching
Chair: Yale Patt

Session 6b: 9:30 - 10:30, Room B
Communication and Multicasts
Chair: Kourosh Gharachorloo

Speculative Execution Via Address Prediction and Data Prefetching
Jose Gonzalez and Antonio Gonzalez
Univ. Politecnica de Catalunya, Barcelona, Spain
Performance Considerations in Software Multicasts
Jörg Cordsen, H.-W. Pohl, and Wolfgang Schröder-Preikschat*
GMD-First Berlin, Germany; *Univ. of Potsdam, Germany
Adaptive Data Prefetching Using Cache Information
Ando Ki, Alan Knowles
Univ. of Manchester, UK
Iteration Space Slicing And Its Application To Communication Optimization
William Pugh and Evan Rosser
Univ. of Maryland, MD, USA
Coffee Break: 10:30 - 11:00
Session 7a: 11:00 - 12:00, Room A
Tree-based and Semi-Structured Applications
Chair: John Merlin

Session 7b: 11:00 - 12:00, Room B
Distributed Shared Memory
Chair: Jim Smith

Compiler And Run-Time Support For Semi-Structured Applications
Nikos Chrisochoides, Induprakas Kodukula, and Keshav Pingali
Cornell University, Ithaca, NY, USA
Design and Performance of the Shasta Distributed Shared Memory Protocol
Daniel J. Scales and Kourosh Gharachorloo
Digital Western Research. Lab., Palo Alto, CA, USA
Conflict-Free Template Access in k-ary and Binomial Trees
Sajal K. Das, Falguni Sarkar, and M.Christina Pinotti*
University of North Texas, Denton, TX, USA; *IEI, Consiglio Nazionale delle Ricerche, Pisa, Italy
An I/O Network Architecture of the Distributed Shared-Memory Massively Parallel Computer JUMP-1
H. Nakajo, S. Ohtani, T. Matsumoto, M. Kohata, K. Hiraki and Y. Kaneda
Kobe Univ., Japan
Lunch Break: 12:00 - 14:00
Session 8a: 14:00 - 15:30, Room A
Compilers
Dennis Gannon

Session 8b: 14:00 - 15:30, Room B
Hardware Features for Performance
Chair: Antonio Gonzalez

Symbolic Evaluation for Parallelizing Compilers
Thomas Fahringer and Bernhard Scholz
Univ. of Vienna, Austria; Vienna Univ. of Technology, Austria
Implementation and Analysis of Path History in Dynamic Branch Prediction Schemes
Shlomo Reches, Shlomo Weiss
Tel Aviv Univ., Israel
A Compiler Algorithm for Optimizing Locality in Loop Nests
Mahmut T. Kandemir, J. Ramanujam, and Alok Choudhary
Syracuse Univ., NY, USA; Louisiana State Univ., Baton Rouge, LA, USA; Northwestern Univ., Evanston, IL, USA
A Victim Cache for Vector Registers
Roger Espasa, Mateo Valero
Univ. Politecnica de Catalunya, Barcelona, Spain
Compile-Time Minimisation of Load Imbalance in Loop Nests
Rizos Sakellariou and John R. Gurd
Univ. of Manchester, UK
The Performance Impact of Exploiting Branch ILP with Tree Representation of ILP Code
Soo-Mook Moon, Kemal Ebcioglu
Seoul Nat. Univ., South Korea; IBM T. J. Watson Research Center, Yorktown Heights, NY, USA
Coffee Break: 15:30 - 16:00
Industrial Session: 16:00 - 18:00, Room A
Chair: Erich Schikuta

Sightseeing Tour and Heurigen Dinner: Departure 19:00 (Penta Hotel)


Friday, July 11

Invited Session: 8:30 - 10:30, Room A
Chair: Jim Smith

Invited Talk: 8:30 - 9:30
The Road to a Petaflops Computer
Paul Smith, Department of Energy, Washington, DC, USA (Speaker)
Martin Sokoloski, Science and Technology Corp., Hampton, VA, USA
Invited Talk: 9:30 - 10:30
SPOCK: Rapid Product Development via Computer-Based Communication
Stefan Rill, Daimler Benz Aerospace Airbus / Univ. of Bremen, Germany
Coffee Break: 10:30 - 11:00
Session 9a: 11:00 - 12:30, Room A
Data Placement and Transformation
Chair: Jörg Cordsen

Session 9b: 11:00 - 12:00, Room B
Performance Prediction and Coding
Chair: Jack Dongarra

Non-singular data transformations: Definition, Validity and Applications
Michael F.P. O'Boyle and P.M.W Knijnenburg
Univ. of Manchester, UK; Leiden Univ., The Netherlands
Developing Architecture Adaptive Algorithms using Simulation with MISS-PVM for Performance Prediction
Dieter F. Kvasnicka and Christoph W. Ueberhuber
Vienna Univ. of Technology, Austria
Cache Miss Equations: An Analytical Representation of Cache Misses
Somnath Ghosh, Margaret Martonosi, and Sharad Malik
Princeton Univ., NJ, USA
Optimizing Matrix Multiply using PHiPAC: a Portable, High-Performance, ANSI C Coding Methodology
Jeff Bilmes, Krste Asanovic, Chee-Whye Chin, and Jim Demmel
Univ. of California at Berkeley, CA, USA
Adaptive Migratory Scheme for Distributed Shared Memory
Jai-Hoon Kim, Nitin H. Vaidya
Texas A&M Univ., College Station, TX, USA

End of Conference: 12:30

Page last modified: July 3, 1997 (Bernd Wender)