CONFERENCES 2925) 8th ACM International Conference on Supercomputing HPCwire ============================================================================= ADVANCE PROGRAM AND REGISTRATION 8th ACM International Conference on Supercomputing Manchester, England, July 11-15, 1994 Sponsored by Association for Computing Machinery/SIGARCH July 11-15, 1994 Ramada Renaissance Hotel, Manchester, England The ACM International Conference on Supercomputing is a forum for engineers and scientists throughout the world to exchange ideas and research results relating to high performance supercomputing systems. The main focus of ICS'94 is high performance and parallel computing systems viewed from three different perspectives: architecture, software and applications. General Chair: John Gurd, University of Manchester Program Chair: William Jalby, Universiti de Versailles Vice Program Chair - Architecture: S. Tomita, Kyoto University Vice Program Chair - Software: M. Wolfe, Oregon Graduate Institute Vice Program Chair - Applications: K. Gallivan, CSRD Program Committee: M.Amamiya, Kyushu University J.-L.Baer, University of Washington D. Callahan, Tera Computer Company E. Davidson, University of Michigan C. Farhat, University of Colorado D. Gannon, Indiana University W. Gentsch, Genias F. Hossfeld, KFA J|lich R. Ibbett, University of Edinburgh M. Ishii, Fujitsu RL Y. Kaneda, Kobe University N. Koike, NEC RL P. Leca, ONERA DMI G. Meurant, CEA Y. Muraoka, Waseda University K. Nakazawa, Tsukuba University N. Nishimukai, Hitachi CRL S. Oyanagi, Toshiba RL Y. Oyanagi, University of Tokyo R. Perrott, Queen's University Belfast C. Polychronopoulos, CSRD T. Priol, Universiti de Rennes A. Rawsthorne, University of Manchester P. Sadayappan, Ohio State University N. Saito, Keio University J. Saltz, University of Maryland T. Shimada, Nagoya University H. Sips, Delft University of Technology J. Smith, Cray Research Inc. M. Valero, UPC Barcelona A. Veidenbaum, CSRD H. Wijshoff, Leiden University Y. Yamaguchi, ETL A. Yonezawa, University of Tokyo T. Yuba, Electro-Communication University H. Zima, University of Vienna Finance Chair: John Sopka, High Performance Computing Engineer Local Arrangements Chair: Brian Warboys, University of Manchester Exhibition Chair: Frank Sumner, University of Manchester FULL TIMETABLE Monday 11th. July A full day tutorial in two parts (see tutorial registration form for further details): 09:30 - 12:30 Parallel Computer Architectures 12:30 - 14:00 Lunch 14:00 - 17:00 Languages and Compilers Michael Wolfe, Oregon Graduate Institute Tuesday 12th. July 09:30 - 10:30 Keynote Speaker 10:30 - 11:00 Break 11:00 - 12:30 Parallel Sessions 1a & 1b 1a: Vector Multiprocessors Distributed Storage Control Unit for the Hitachi S-3800 Multivector Supercomputer: K. Kitai, T. Isobe, T. Sakakibara, S. Yazawa, Y. Tamaki and T. Tanaka and K. Ishii A Model for Dataflow Based Vector Execution: W. M. Miller, W. A. Najjar, A. P. W. Bvhm Synchronized Access to Streams in SIMD Vector Multiprocessors: M. Peiron, M. Valero and E. Ayguadi 1b: Parallelizing Compilers The PRIVATIZING DOALL Test: A Run-Time Technique for DOALL Loop Identifica- tion and Array Privatization: L. Rauchwerger and D. Padua Reducing Data Communication Overhead for DOACROSS Loop Nests: P. Tang and J. N. Zigman Evaluating Automatic Parallelization for Efficient Execution on Shared Memory Multiprocessors: K. S. McKinley 12:30 - 14:00 Lunch 14:00 - 15:00 Parallel Sessions 2a & 2b 2a: Cache Coherence I An Evaluation of Directory Protocols for Medium-Scale Shared-Memory Multiprocessors: S. S. Mukherjee and M. D. Hill An Evaluation of a Compiler Optimization for Improving the Performance of a Coherence Directory: F. Mounes-Toussi, D. L. Lilja and Z. Li 2b: Applications I Parallelisation of the SDEM Distinct Element Stress Analysis Code on the KSR1: G. K. Egan, G. D. Riley and J. M. Bull Ultrasonic Wave Propagation on Parallel Machines: J.-P. Grigoire, C. Domain and B. Thomas 15:00 - 15:30 Break 15:30 - 17:00 Parallel Sessions 3a & 3b 3a: Compiler Optimization An Efficient Approach to Computing Fix-points for Complex Program Analysis: L.-L. Chen and W. L. Harrison III Optimal Local Register Allocation for a Multiple-Issue Machine: W. M. Meleis and E. S. Davidson Scheduling Reductions: X. Redon and P. Feautrier 3b: Networks A Dominating Set Model for Broadcast in All-Port Wormhole-Routed 2D Mesh Net-works: Y. Tsai and P. K. McKinley The Interaction between Virtual Channel Flow Control and Adaptive Routing in Wormhole Networks: S. Ramany and D. Eager Fault-Tolerant Wormhole Routing in Tori: S. Chalasani and R. V. Boppana 17:00 - 18:00 Break 18:00 - 20:00 Reception and Buffet 20:00 - 21:30 Mini-symposia Wednesday 13th. July 09:30 - 10:30 Invited Speaker 10:30 - 11:00 Break 11:00 - 12:30 Parallel Sessions 4a & 4b 4a: Performance Evaluation I Performance of the CM-5 Scalable File System: T. T. Kwan and D. A. Reed Communication in the KSR1 MPP: Performance Evaluation Using Synthetic Work- load Experiments: E. L. Boyd and E. S. Davidson Architecture Implications of High-Speed I/O for Distributed-Memory Computers: T. Gross and P. Steenkiste 4b: Scheduling Combining Static and Dynamic Scheduling on Distributed-Memory Multiprocessors: O. Plata An Optimal Upper Bound on the Minimal Completion Time in Distributed Supercomputing: L. Lundberg and H. Lennerstad Compiler Techniques for Maximizing Fine-grain and Coarse-grain Parallelism in Loops with Uniform Dependences: Y.S. Chen, S.-D. Wang and C.-M. Wang 12:30 - 14:00 Lunch 14:00 - 17:00 Vendor Session and Vendor Panel 17:00 - 19:00 Vendor Reception Thursday 14th. July 09:30 - 10:30 Invited Speaker 10:30 - 11:00 Break 11:00 - 12:30 Parallel Sessions 5a & 5b 5a: Analysis of Parallel Irregular Computations Data and Program Restructuring of Irregular Applications for Cache-Coherent Multi-processors: K. A. Tomko and S. G. Abraham Nonzero Structure Analysis: A.J.C. Bik and H.A.G. Wijshoff Techniques to Overlap Computation and Communication in Irregular Iterative Applications: A. Lain and P. Banerjee 5b: Networks II Throughput Performance of a Synchronous, Circuit-switched Interconnection Cached Network for Multicomputers: V. Gupta and E. Schenfeld Nonblocking Multirate Broadcast Networks for High-performance Computing Systems: Y. Yang 12:30 - 14:00 Lunch 14:00 - 15:00 Parallel Sessions 6a & 6b 6a: Cache Coherence II Exploiting Cache Affinity in Software Cache Coherence: H. Li and K. C. Sevcik Performance Evaluation of Hybrid Hardware and Software Distributed Shared Memory Protocols: R. Chandra, K. Gharachorloo, V. Soundararajan and A. Gupta 6b: Applications II Limited Area Numerical Weather Forecasting on a Massively Parallel Computer: G. Cats, N. Gustafsson and L. Wolters Fast Phylogenetic Analysis on a Massively Parallel Machine: H. Matsuda, G. J. Olsen, R. Overbeek and Y. Kaneda 15:00 - 15:30 Break 15:30 - 17:00 Panel Session 17:00 - 19:00 Break 19:00 - 21:00 Banquet Friday 15th. July 09:30 - 10:30 Invited Speaker 10:30 - 11:00 Break 11:00 - 12:30 Parallel Sessions 7a & 7b 7a: Performance Evaluation II Compiling Performance Models from Parallel Programs: A.J.C. van Gemund Queueing Performance Analysis of CoScheduling in a Pool of Processors Envi- ronment: M.A. Schaar, K. Efe and W. Shang Normalized Performance Indices for Message Passing Parallel Programs: S.R. Sarukkai, J. Yan and J. K. Gotwals 7b: Data Locality Exploitation The Effectiveness of Caches for Vector Processors: J. D. Gee and A. J. Smith Using Virtual Lines to Enhance Locality Exploitation: O. Temam and Y. Jegou MOB Forms: A Class of Multilevel Block Algorithms for Dense Linear Algebra Operations: J. J. Navarro, A. Juan and T. Lang 12:30 - 14:00 Lunch 14:00 - 16:00 Parallel Sessions 8a & 8b 8a: Data Distribution An Approach to Communication-Efficient Data Redistribution: S. D. Kaushik, C.-H. Huang, R. W. Johnson and P. Sadayappan Twisted Data Layout: T. Shindo, H. Iwashita and S. Kaneshiro Compiler and Runtime Support for Out-of-Core HPF Programs: R. Thakur, R. Bordawekar and A. Choudhary Compiling Block-Cyclic Distributions in Fortran D: S. Hiranandani, K. Kennedy, J. Mellor-Crummey and A. Sethi 8b: Sparse Matrix Computations A Generalized Vision of Some Parallel Bidiagonal Systems Solvers: J.-L. Larriba-Pey, J. J. Navarro, O. Roig and A. Jorba Analysis of Non-strict Functional Implementations of the Dongarra-Sorensen Eigensolver: S. Sur and W. Bvhm The Parallel Solution of Unsymmetric Sparse Linear Systems Using the H^* Reordering and an Associated Factorization: K.A. Gallivan, B.A. Marsolf and H.A.G. Wijshoff On Sparse Matrix Reordering for Parallel Factorization: B. Kumar, P. Sadayappan and C.-H. Huang 16:00 - 16:30 Close REGISTRATION AND ACCOMMODATION DETAILS Please register your wish to attend ICS 94 and/or the tutorial day by completing the enclosed resistration form(s). Please note the two alternative sections which need to be completed depending on whether you are registering in the United States or the United Kingdom. The main conference registration fee does not include the fee for the tutorial day on Monday 11th. July, or the conference banquet. Lunches are, however, included. Accommodation has been reserved for delegates at preferential rates at the Ramada Renaissance Hotel and details can be found on the registration form. More economical accommodation can be provided in University Halls of Residence, but these are located 1 to 2 miles from the Ramada Renaissance Hotel. Details of this accommodation will be provided on request. A number of excursions have also been planned and details of these will be forwarded on receipt of the completed registration form. Please note that you should quote your ACM Membership Number to qualify for the ACM Member Rate and that the presentation of student identification will be required at the conference to qualify for the student rate. ********************************************************************** MAIN CONFERENCE REGISTRATION FORM 8th ACM International Conference on Supercomputing Manchester, England, July 11-15, 1994 To register, please complete this form and send it with payment to the address shown overleaf: Name: ACM Membership No. Address: Country Postal Code Phone: Fax: Email: Main Conference Registration Fee (Please tick appropriate box): This fee includes all lunches and refreshments during the conference. Advance Registration Late Registration (Received by May 31st 1994) (Received after June 1st. 1994) o ACM Member #260.00 ($390.00) o ACM Member #325.00 ($490.00) o Non-Member #330.00 ($495.00) o Non-Member #395.00 ($595.00) o Student #165.00 ($250.00) o Student #195.00 ($295.00) Conference Banquet Registration Fee o #30.00 ($45.00) The conference banquet is not included in the main registration fee. Accommodation A number of rooms have been reserved for delegates at the RamadaRenaissance Hotel at specially discounted rates for ICS 94. Reservations should be made by completing the section below and adding the appropriate amount to the registration fee. The full amount should be sent to the Registration Chair (see payment details overleaf). Bed & Breakfast Sun. 10th Mon. 11th Tues. 12th Weds. 13th Thurs. 14th Single Room #60.00 ($90.00) per night Twin/Double sharing #40.00 ($60.00) per night If you are sharing a room, please indicate here if there is a particular delegate with whom you would like to share. Total cost of accommodation # ($) Cancellation Registration cancellations must be received by the UK Registration Chair at the address shown overleaf. If a refund is due, it will be made after the conference. Registration cancellation by May 31st 1994: 60% refund Registration cancellation after May 31st 1994: No refund Accommodation cancellations must also be received by the UK Registration Chair. Again, if a refund is due, it will be made after the conference. UNITED STATES REGISTRATION Please send your completed application form with payment to: John Sopka (US Registration Chair) Tel: 1-412-561-3057 PO BOX 2576 Email: sopka@eclipse.pgh.pa.us Pittsburgh, PA 15230 United States Payment Method (Must be in US $) o Credit Card o Visa o Mastercard o American Express Card Number Expiry Date Cardholder's Name Signature o Cheque payable to ICS 94 o Bank Transfer through the to: (Name of your bank) Account Name: ICS 94 Account Number: 852-07N19 Bank ID: 1011730 Branch: Mellon Bank 1 Mellon Bank Plaza Pittsburgh, PA 15258-0001 United States Please attach a copy of your Bank Transfer Order. UNITED KINGDOM REGISTRATION Please send your completed application form with payment to: Ursula Hayes (UK Registration Chair) Tel: +44 (0)61 275 6172 Department of Computer Science Fax: +44 (0)61 275 6200 University of Manchester Email: ursula@cs.man.ac.uk Oxford Road Manchester M13 9PL England Payment Method (Must be in UK Pounds) o Cheque payable to The University of Manchester o Bank Transfer through the to: (Name of your bank) Account Name:Victoria University of Manchester Account Number: 026-11-813 Bank Sort Code: 01-05-31 Branch: National Westminster Bank 35 King Street Loyd Entwisle Street Manchester, England Please attach a copy of your Bank Transfer Order. ************************************************************************** TUTORIAL REGISTRATION FORM 8th ACM International Conference on Supercomputing Manchester, England, July 11-15, 1994 To register for the tutorial day on Monday 11th. July, please complete this form and send it with payment to the address shown overleaf: Name: ACM Membership No. Address: Country Postal Code Phone: Fax: Email: Tutorial Day Registration Fee o #120.00 ($180.00) Monday July 11th. 1994 Parallel Computer Architectures, Languages and Compilers A full day tutorial in two parts. Parallel computer architectures have been designed and used for several decades. Many forms of parallelism are now used in commercial as well as research architectures. Effective use of parallel computers depends on the use of appropriate algorithms, languages and compilers. The first part of this tutorial covers the spectrum of parallelism as used in modern computer architectures. The material focuses on parallel architectures designed to enhance the performance of scientific and engineering applications. The range of parallelism ranges from instruction level parallelism to massively parallel multiprocessors. The performance of these machines depends to a great extent on the quality of compiled code; thus we will also discuss algorithmic, language and compiler issues related to the various architectural styles. The material in this part is aimed at a general audience with some knowledge of parallel computing. The second part of the tutorial will focus on advanced compiler issues for parallelism. The topics will include recent advances in program analysis and transformations. A set of tools will be developed, followed by discussion of how to apply the tools for various parallel architectures, e.g., SIMD, MIMD, shared-memory, message-passing, etc. The material will include compiling sequential programs for parallel machines, as well as analysis and transformation of parallel languages. The material in this part is aimed at an audience with some knowledge of parallel languages and computer architectures; knowledge of the material in the first part of this tutorial is sufficient. Registration will include a copy of the instructional materials. Instructor: Michael Wolfe Associate Professor of Computer Science and Engineering Oregon Graduate Institute of Science and Technology UNITED STATES REGISTRATION Please send your completed application form with payment to: John Sopka (US Registration Chair) Tel: 1-412-561-3057 PO BOX 2576 Email: sopka@eclipse.pgh.pa.us Pittsburgh PA 15230 United States Payment Method (Must be in US $) o Credit Card o Visa o Mastercard o American Express Card Number Expiry Date Cardholder's Name Signature o Cheque payable to ICS 94 o Bank Transfer through the to: (Name of your bank) Account Name: ICS 94 Account Number: 852-07N19 Bank ID: 1011730 Branch: Mellon Bank 1 Mellon Bank Plaza Pittsburgh, PA 15258-0001 United States Please attach a copy of your Bank Transfer Order. UNITED KINGDOM REGISTRATION Please send your completed application form with payment to: Ursula Hayes (UK Registration Chair) Tel: +44 (0)61 275 6172 Department of Computer Science Fax: +44 (0)61 275 6200 University of Manchester Email: ursula@cs.man.ac.uk Oxford Road Manchester M13 9PL England Payment Method (Must be in UK Pounds) o Cheque payable to The University of Manchester o Bank Transfer through the to: (Name of your bank) Account Name:Victoria University of Manchester Account Number: 026-11-813 Bank Sort Code: 01-05-31 Branch: National Westminster Bank 35 King Street Loyd Entwisle Street Manchester, England Please attach a copy of your Bank Transfer Order. ***************************************************************************** H P C W I R E S P O N S O R S Product specifications and company information in this section are available to both subscribers and non-subscribers. 901) ANS 902) IBM Corp. 904) Intel SSD 905) Maximum Strategy 906) nCUBE 907) Digital Equipment 909) Fujitsu America 912) Avalon Computer 914) Applied Parallel Res. 915) Genias Software 916) MasPar Computer 921) Cray Research Inc. 927) ISR Corp. ***************************************************************************** Copyright 1994 HPCwire. To receive the weekly HPCwire at no charge, send e-mail without text to "trial@hpcwire.ans.net".