NetNews Usenet Archive 1992 #27

home *** CD-ROM | disk | FTP | other *** search

/ NetNews Usenet Archive 1992 #27 / NN_1992_27.iso / spool / comp / parallel / 2528 < prev next >

Wrap

Text File | 1992-11-16 | 14.9 KB | 498 lines

Newsgroups: comp.parallel Path: sparky!uunet!think.com!rpi!gatech!hubcap!fpst From: dduke@ibm1.scri.fsu.edu (Dennis Duke) Subject: Cluster Computing Workshop - Tentative Schedule of Speakers Message-ID: <1992Nov16.131718.4765@hubcap.clemson.edu> Sender: news@sun13.scri.fsu.edu Organization: SCRI, Florida State University Date: 13 Nov 92 21:30:10 GMT Approved: parallel@hubcap.clemson.edu Lines: 486 WORKSHOP ON CLUSTER COMPUTING (Tentative Schedule) Supercomputer Computations Research Institute Florida State University Tallahassee, Florida 32306 Wednesday, December 2 - Friday, December 4, 1992 Organizers: Peter Dragovitsch (drago@scri.fsu.edu), SCRI, FSU Dennis Duke (dduke@scri.fsu.edu), SCRI, FSU Uli Hansmann (hansmann@scri.fsu.edu), SCRI, FSU Neil Lincoln (nrl@nips.ssesco.com), SSESCO Timothy Mattson (mattson@sca.com), Yale and SCA Vaidy Sunderam (vss@mathcs.emory.edu), Emory University The goal of the workshop is to bring together people who are interested in the issues of using heterogeneous clusters of computers as computational resources. This group of people would include: - computational scientists or other end-users who desire to exploit the power and capabilities of heterogeneous clusters as an alternative to conventional supercomputers or MPP's - software developers of queueing systems, parallel processing packages, and other software tools - system administrators interested in both the strategic and technical issues of running clusters, including in general any kind of heterogeneous collection of networked computers - vendors who are developing products to serve this market: (a) new generations of high performance workstations (b) new forms of packaging of workstations (c) new high speed networking products (d) new products for addressing mass storage and other I/O needs (e) any other applicable technology - persons who would like to share their experiences, and especially give critical advice on desired user requirements, shortcomings of present hardware and software configurations, successes and failures to date, overviews of planned projects, etc. Although most practical implementations of clusters to date consist of collections of RISC workstations, we are interested also in more general configurations which might include any number of distinct architectures, and a variety of new high-speed network interconnections. The format of the workshop will be to fill the day with contributed and invited talks, and to have informal evening gatherings designed to further enhance the opportunity for information exchange. We especially encourage contributions of a 'practical experience' nature, since this is likely to be of the most use to the many people who are involved in cluster computing, and will also be complimentary to the many conferences that stress more the academic side of computing and computer/computational science research. The agenda for the workshop will be as follows: Tuesday, December 1 6:00 - 8:00 pm Registration and hosted Reception - Holiday Inn Wednesday, December 2 7:30 - 8:30 Continental Breakfast and Registration - Conference Center 8:30 - 9:15 Tilak Agerwala, IBM Kingston 9:15 - 10:00 Dan Anderson, N.C.A.R. The care and feeding of the NCAR cluster 10:00 - 10:30 Break 10:30 - 12:00 Session 1 Joe Mathis, IBM Austin Fibre Channel Standard - Architecture Terry Anderson, Ancor Communications Two-Dimensional Networking Jim Feeney, IBM Endicott A Prototype Switch for Scalable High Performance Distributed Computing 10:30 - 12:00 Session 2 Richard P. LaRowe Jr., Worcester Polytechnic Institute Heterogeneous Computing Clusters By Design Doug Elias, Cornell Distributed Applications Management (DAM): Adaptive mappings between distributed applications and network computing resources. Joshua Auerbach, IBM T. J. Watson Research Center Concert/C: C language extensions for convenient distributed programming 12:00 - 1:00 Lunch 1:00 - 3:00 Rusty Lusk, ANL Portable Parallel Computing with p4 Tim Mattson, Yale and SCA Computing with LINDA Adam Kolawa, Parasoft Computing with EXPRESS Vaidy Sunderam, Emory Computing with PVM 3:00 - 3:30 Break 3:30 - 5:30 Session 3 Josef Fritscher, Technical University of Vienna Just Another Parallel Extension for C++ Martin Rinard Heterogeneous Parallel Programming in Jade Juan Leon, Allan Fisher, Peter Steenkiste, CMU Fail-safe PVM Eugene C. Boman and Angela B. Mounts United Technologies Research Center Practical Experience with PVM and CFD 3:30 - 5:30 Session 4 Andrew S. Grimshaw, Jon Weissman, Emily A. West, UVa Experiences with the Mentat MetaSystems Testbed R. Jagannathan, SRI International Parallel and Fault-Tolerant Processing on Workstation Clusters Yuan Shi, Temple University The Synergy Infrastructure - Tools for Computing the Future Ronald L. Rockhold, IBM Austin Developing and Testing Operating Systems on non-Cache Coherent Multiprocessors 7:00 - Hosted Reception at SCRI 7:00 - 8:00 Demos at SCRI 8:00 - 9:30 Moderated Session Parallel/Distributed Computing environments - "Is There Common Ground? Robert Voigt, NSF, moderator Thursday, December 3 7:30 - 8:30 Continental Breakfast and Registration - Conference Center 8:30 - 9:15 Jack Dongarra, ORNL and University of Tennessee LAPACK for Clusters 9:15 - 10:00 Martin Schultz, Yale Solving the Shallow Water Equations on Clusters 10:00 - 10:30 Break 10:30 - 12:00 Session 5 Patrick T. Homer, Richard D. Schlichtin, University of Arizona Using the Schooner Interconnection System for Constructing Scientific Applications from Heterogeneous Resource Carolyn Gard, UGa UGA Experiences with Clustering RS/6000s and the ES/9000 Steve Wolbers, Frank Rinaldo, Fermilab The Farms Systems at Fermilab 10:30 - 12:00 Session 6 Carl Albing, Cray Research Load Distribution using NQS Maria Tortorici, Marco Alfano and G. Lo Re Centro Universitario di Calcolo - University of Palermo Distributing Numerical Algorithms: Some Experiences with NCS and PVM Songnian Zhou, University of Toronto and Platform Computing Corporation LSF: Load Sharing in Large-Scale Heterogeneous Distributed Systems 12:00 - 1:00 Lunch 1:00 - 3:00 Session 7 Miron Livny, University of Wisconsin-Madison What Did We Learn From Serving a Quarter of a Million Batch Jobs on a Cluster of Privately Owned Workstations Beverley Pederson, IBM Research, Yorktown Heights Condor Distributed Scheduler Steven Bellenot, FSU A Network Version of the Time Warp Operating System Alan G. Yoder, Arindam Banerji, John M. Tracey, David L. Cohn University of Notre Dame Clustered Computing Research at Notre Dame 1:00 - 3:00 Session 8 Harold Trease, Jim Tabor, LANL Optimal Mapping of Unstructured Meshes to MPP and Clusters Using Recursive Spectral Bisection Stephen Fleischman, Convex Computer Corporation Some Experiences with Molecular Dynamics on an HP Cluster Jose M. Sanz, NASA Lewis Research Center A Clustered Cumputer Platform for the Inverse Design of a Full Engine Michael Pernice, Utah Supercomputing Institute Network Parallel Solution of Nonsymmetric PDEs: How Well Can RISC Clusters Compete with Distributed Memory Multiprocessors? 3:00 - 3:30 Break 3:30 - 5:30 Session 9 Martin Graff, Vienna University Computer Center Vienna Queueing System (VQS) L. Busby, LLNL Cluster Experiences at LLNL Mark Swanson, University of Utah The Schizophrenic Workstation Jim Hudgens, SCRI, FSU DQS 3:30 - 5:30 Session 10 Dan Nessett and Jim Rathkopf, LLNL Computing on Heterogeneous Supercomputer Clusters Salvatore Filippone and Maria Loredana Sales, IBM ECSEC, Rome Parallel Linear Systems Solvers on RISC/System~6000 Clusters R.C. Armstrong, R.E. Cline, Jr., M.L. Kozykowski, Sandia National Laboratories and J.F. Macfarlane, Lawrence Berkeley Laboratory A Toolkit Approach to Distributed Computing Ronald G. Minnich, The Supercomputing Research Center A SPARCStation ELC-based cluster 7:00 - Hosted Reception at SCRI 7:00 - 8:00 Demos at SCRI 8:00 - 9:30 Moderated Session Public Domain versus Commercial Software Neil Lincoln, SSESCO, moderator Friday, December 4 7:30 - 8:30 Continental Breakfast - Conference Center 8:30 - 10:00 Session 11 K. W. Johnson, J. Bauer, G. A. Riccardi, FSU and K. K. Droegemeier, University of Oklahoma Distributed processing of a regional prediction model Craig J. Tremback, Robert L. Walko, and William R. Cotton Colorado State University CONSIDERATIONS FOR THE PARALLELIZATION OF AN ATMOSPHERIC NUMERICAL MODEL to be determined 8:30 - 10:00 Session 12 Lewis Patterson, University of Alabama Birmingham Construction of a Fault-Tolerant Distributed Tuple-Space Bob Hyatt, University of Alabama Birmingham Future Developments in Tuple-Space Roberto Gomperts, Rosario Caltabiano, Silicon Graphics A Simple Approach To Network Parallelism 10:00 - 10:30 Break 10:30 - 12:00 Session 13 Susan Ying and Madhavan Narayanan, SCRI, FSU SOLVING THE NAVIER-STOKES EQUATIONS IN HETEROGENEOUS NETWORKS Paul M. Beaumont and Lili Yuan, SCRI, FSU Optimization Using Parallel Genetic Algorithms: Applications in Economics to be determined 10:30 - 12:00 Session 14 Stephen Taylor, California Institute of Technology Composing Concurrent Programs Saul Youssef and Phillip Rulon, SCRI, FSU Coarse Grained Parallel Function Evaluation Randy Langley, SCRI, FSU DSM: An Explicit Distributed Shared Memory Libary 12:00 Workshop Ends It is generally possible to leave Tallahassee in the afternoon and reach most destinations on either the east or west coast by Friday evening. Among the institutions which will be represented at the Workshop are: Alcoa Technical Center Ancor Communications Applied Parallel Research Arco Oil Company Argonne National Laboratory Boeing Company Brookhaven National Laboratory Caltech Carnegie Melon University Chevron Oil Company CICG (France) CNUCE Institute of CNR Pisa (Italy) Colorado State University Convex Computer Corporation Cornell Theory Center Cray Research Danish Computing Center (Denmark) Emory University Fermilab Florida State University George Washington University IBM Austin IBM Endicott IBM European Center for Scientific and Engineering Computing (Italy) IBM Kingston IBM T J Watson Research Center Lawrence Berkeley Laboratory Lawrence Livermore National Laboratory Los Alamos National Laboratory Lulea University of Technology (Sweden) NASA Lewis Research Center National Center for Atmospheric Research Naval Research and Development Center Oak Ridge National Laboratory Parasoft Corporation Pratt & Whitney Aircraft SSESCO Sandia National Laboratories Schlumberger/Geco-Prakla Scientific Computing Associates Silicon Graphics SRI International Stanford University Sterling Software Syracuse University Technical University of Vienna (Austria) Temple University United Technologies Research Center University of Alabama Birmingham University of Arizona University of Georgia University of Houston University of Kentucky University of Illinois University of Notre Dame University of Nova Scotia University of Palermo (Italy) University of Tennessee University of Toronto (Canada) University of Utah University of Vienna (Austria) University of Virginia University of Wisconsin Utah Supercomputing Institute Please register for the workshop using the form attached below. We encourage as many attendees as possible to plan to make a presentation at the workshop. We do ask that speakers plan to address topics within the scope outlined above. Please send us a short abstract describing your talk, so we can plan an appropriate place in the schedule. An announcement of the schedule of speakers will be distributed as soon as possible. Within the limits of available time, we would like to accomodate as many speakers as practicable. The 'proceedings' of the workshop will be published via anonymous ftp. We will request each speaker to send us an approriate electronic version of his talk (ascii, postscript, tex, latex, troff, etc.). These will then be placed on the machine ftp.scri.fsu.edu for further distribution. Any questions or requests can be sent via email to cluster-workshop@scri.fsu.edu or to one of the organizers. SCRI can be reached by phone at (904)-644-1010. =============================================================================== REGISTRATION AND HOTEL INFORMATION WORKSHOP ON RISC CLUSTER COMPUTING December 2-4, 1992 PLEASE TYPE OR PRINT Name _____________________________________ Social Security Number ___________________ Company __________________________________ Address/Mailstop __________________________________________________ City/State/Zip/Country ____________________________________________ Email address _______________________________________ Phone ( )______________________ Workshop Program Number: 1903293 The registration fee for the workshop is $85, and includes three continental breakfasts, Wednesday and Thursday lunches, morning and afternoon break refreshments, and the food and drink for the evening sessions. There will be no workshop dinners. If you want to pay by check, please print out and fill in the above registration form, including especially the Workshop Program Number, and return the registration form and fee to: The Registrar Florida State University Conference Center 555 West Pensacola Street Tallahassee, Florida 32306 If you want to register by credit card, you may register by email by sending back an edited form of the above registration information, as well as the following credit card information: Credit Card Name _____________________________Visa or Mastercard only Credit Card Number ___________________________ Name (as it appears on card) __________________________________ Expirations Date of Card _____________________ There is an additional 2% charge by the University for credit card registrations (bringing the total to $86.70). Email registrations, and all other inquiries about the workshop, may be sent to: email: cluster-workshop@scri.fsu.edu, or fax: (904)-644-0098, attention of Pat Meredith Hotel Information: The workshop hotel is the Holiday Inn, located within walking distance of the FSU Conference Center and SCRI on the FSU campus. The hotel does provide shuttle service to/from the Tallahassee airport. Rooms are $39 per night, single or double, and reservations should be made by November 10. Be sure to mention the Workshop on RISC Cluster Computing to get the special rate. Holiday Inn/Downtown University Center 316 West Tennessee Street Phone: (904)-222-8000 Tallahassee, FL 32301 Fax: (904)-222-8113