NERSCPowering Scientific Discovery Since 1974


ERSUG Meeting

July 11-12, 1994

Holiday Inn, Crowne Plaza

Rockville, MD

The Energy Research Supercomputer Users' Group (ERSUG) will meet at the Holiday Inn, Crowne Plaza in Rockville, MD on July 11-12, 1994. In the past, this meeting has combined presentations describing work-in-progress at NERSC with lively user discussions in the areas of the services and capabilities provided by NERSC. For this particular meeting, however, the focus will change somewhat. First, more emphasis will be put on future services, and second, the ERSUG meeting will be combined with the Energy Research Power Users Symposium (ERPUS). The session dedicated to ERPUS will occur on the second day of the two day ERSUG meeting and will be devoted to user presentations describing results obtained through the Special Parallel Processing (SPP) program at NERSC, as well as results obtained through use of other DOE high-performance systems.

The general theme of the first day of the ERSUG meeting will be the realized and projected improvements in the NERSC computing environment as well as on some remaining bottlenecks inhibiting most efficient access to system resources. A report from the committee to study options for controlling disk space usage will be part of a discussion of integrating host disk systems with AFS and archival storage at NERSC. There will be a progress report from ESNet and a detailed report on the transition to X-windows as a baseline standard for our users as we move into an era in which vendor software is driven by, and only compatible with, X-based standards.

Also on the first day, a new program initiated by the Office of Scientific Computing to arrange for Massively Parallel (MP) computing access will be described. NERSC Principal Investigators are now applying for access to parallel computers at the High Performance Research Supercomputer Centers at Los Alamos and Oak Ridge and to the Livermore T3D. Such access will allow scientists to develop applications in anticipation of the arrival of the NERSC MP system. Following this will be an update on the status of the NERSC procurement and a discussion of the anticipated impact that the transition will have on our users. This will include an approach for addressing the capacity-capability antithesis, not only during the transition period, but also beyond.

Since the meeting is dominated by what is about to happen rather than by what has happened, the NERSC staff will prepare a number of short documents describing recent progress and the status of critical areas. These will have to substitute for detailed presentations and will include topics such as the Energy Research Decision Package (ERDP), Centralized User Bank (CUB), etc.

We expect this meeting to be very stimulating; its focus on the present will be on extracting the maximum out of existing resources, and the focus on the future will be on planning a feasible and realistic path to effective use of new technologies. In this journey, the major constraints will be to find a reasonable path for all of our users regardless of their computing requirements, since good science is not always the biggest science...


            Agenda for the ERSUG meeting and ERPUS session


                             Monday, July 11, 1994


 8:30     Welcoming Remarks - Jack Byers

 8:40   Welcome from Washington - John Cavallini

 8:45     Washington View - Tom Kitchens

 9:05     Production Systems Plans - Moe Jette and Rick Kendall

                New storage usage paradigms

                Disk reconfiguration committee report

                CUB plans

10:05     BREAK

10:15     ESnet Plans - Jim Leighton

          Backbone upgrade

          Network services

11:00     X-windows: How To Get From Here To There - Barry Howard and

                                                   Alice Koniges

12:00     LUNCH (ExERSUG meeting)

 1:30     SPP Project Status and Plans for 1995 - Bruce Curtis and Tom Kitchens,

                                                Jack Byers

          Who are the participants?

          A survey of current attained efficiencies

                SPP in 1995

                Discussion of capacity and capability: the balance in 1995

 2:15     MPP Procurement Status - Michel McCoy

 2:45     BREAK

 3:00     Transition to Parallel Computing - Tammy Welcome and Steve Louis

          Access to the HPCRCs and to the LLNL T3D

          Capacity and capability solutions for 1996 and beyond

          Storage expands to meet demands

 4:00   R and D: Tracking the Computational Explosion - Alice Koniges

 4:20   Open Discussion

 5:00     Adjourn


                             Tuesday, July 12, 1994

 8:30     NERSC's Role as a High Capability Computing Resource - Bill McCurdy


                                 ERPUS Session

    User presentations describing results obtained through the Special Parallel

    Processing (SPP) program at NERSC, as well as results obtained through use

    of other DOE high-performance systems

 9:20  "Recent Progress on Lattice QCD with MIMD Parallel Computers"

     Claude Bernard, Washington University       

 9:40  "Understanding Tokamak Gyrofluid Turbulence Through Visualization"

        Gary Kerbel, NERSC

10:00  "Parallelization of Quantum Chemistry Codes" - Al Wagner, ANL

10:20   BREAK

10:40  "Global Ocean Modeling" - Bob Malone, LANL

11:00  "Adaptive Methods for Computational Fluid Dynamics"

        Philip Colella, UC Berkeley

11:20  "Recent Results of Lattice QCD Simulations"

     Stephen Sharpe, University of Washington

11:40  "Tokamak Plasma Turbulence Simulations" - Greg Hammett, PPPL

12:00   LUNCH

13:20  "Electronic Structure of Organic Superconductors" - Dale Koelling, ANL

13:40  "Benchmark Ab Initio Molecular Dynamics Simulations of Partially

     Solvated Cations" - Rick A. Kendall, PNL

14:00  "Molecular Views of Damaged DNA:  Adaptation of the Program DUPLEX to

     Parallel Architectures" - Brian Hingerty, ORNL

14:20  "High Performance Computing for Beam Physics Applications"

     Robert Ryne, LANL

14:40  "Parallel Plasma Fluid Turbulence Calculations" - Jean-Noel Leboeuf, ORNL

15:00  "Nuclear Structure from Lattice QCD" - K.-F. Liu, Univ. of Kentucky


                            ERPUS PRESENTERS


Claude Bernard      

Department of Physics, CB1105

Washington University

St. Louis, MI      Phone:  (314)-935-6280

                                 (805)-893-2742 (until 6/28/94)



Philip Colella

Department of Mechanical Engineering

Univ. of California/Berkeley



Greg Hammett

Princeton Plasma Physics Laboratory

P. O. Box 451

Princeton, NJ  08543    Phone: (609)-243-2495



Brian Hingerty

Health Sciences Res. Div.

Oak Ridge National Laboratory

P.O. Box 2009

Mail Stop 8077, Oak Ridge, TN  37831        Phone: (615)-574-0844



Rick A. Kendall

Mail Stop K1-90

High Performance Computational Chemistry Group

Pacific Northwest Laboratory

Richland, WA 99352-0999 (Rick A Kendall)    Phone: (509)-375-2602



Gary Kerbel

Lawrence Livermore National Laboratory


Livermore, CA  94550      Phone: (510)-422-4227



Dale Koelling


Argonne National Laboratory

9700 South Cass Avenue

Argonne, IL  60439      Phone:  (708)-252-5507

   After August 1, 1994:

     Division of Materials Sciences, ER-132, U.S. Department of Energy

     Washington, DC  20585   Phone:  (301)-903-3426



Jean-Noel Leboeuf

Fusion Energy Div.

Oak Ridge National Laboratory

P.O. Box 2009

Mail Stop 8071, Oak Ridge, TN 37831   Phone: (615)-574-1127



K.-F. Liu

Department of Physics and Astronomy

Univ. of Kentucky

Lexington, KY 40506          Phone: (606)-257-4849



Bob Malone

Los Alamos National Laboratory

ACL Mail Stop B287

Los Alamos, NM 87545      Phone: (505)-665-4530



Robert Ryne

Mail Stop H817

Los Alamos National Laboratory

Los Alamos, NM  87545                Phone: (505)-667-8111



Stephen Sharpe

Department of Physics

University of Washington

Seattle, WA



Al Wagner

Argonne National Laboratory

Bldg 200 R105

Argonne, IL 60439        Phone: (708)-252-3597