Agenda
ERSUG Meeting June 13 & 14, 1995
Lawrence Livermore National Laboratory
National Energy Research Supercomputer Center (NERSC)
Bldg. 451, Room 1025 (White Room)
Portions of the ERSUG meeting will be broadcast on the MBONE.
Tuesday, June 13, 1995
8:30 Welcoming Remarks - Brian Hingerty
8:45 Greetings from Tom Kitchens
8:50 Welcome from NERSC - Bill McCurdy
Review of the Agenda
****STRATEGIC PLANNING ISSUES****
9:10 Strategic Planning at NERSC - Bill McCurdy, Michel McCoy
Developments Planned for 1995-1996
Status of the NERSC procurement
10:00 A User Requirements Document (Greenbook) for the Late 90's
- Brian Hingerty
10:15 Break
****PLANNING FOR THE NEXT GENERATION OF HIGH PERFORMANCE SYSTEMS ****
10:30 National Storage Laboratory (NSL) Technology Base System
Progress Report - Keith Fitzgerald
10:45 MPP Update (MPP Access Program, MPP Workshop) - Tammy Welcome
11:10 The Continuing Experiment with Symmetric Multi-Processors (SMPs):
The SGI Power Challenge - Betsy Foote
11:25 Lunch (ExERSUG working lunch in the White Room)
- ExERSUG Functions - Should We Model Ourselves After ESnet/DCCC?
- ExERSUG Membership
****DISCUSSIONS WITH HEADQUARTERS****
1:00 ERSUG Discussion with SAC members via Televideo Conference
Washington View - Tom Kitchens
-Comments Regarding the Recompetition of NERSC
Discussion with SAC members regarding user requirements for
computing
2:45 Break
**PLANNING FOR THE NEXT GENERATION OF HIGH PERFORMANCE SYSTEMS (Cont) ****
3:00 Status and Directions for Graphics at NERSC - Chris Andersonw
3:20 Utilizing a Network of Workstations to Facilitate Supercomputing
-John Allen
****PRODUCTION ENVIRONMENT PROGRESS REPORT****
3:35 Unified Production Environment Update - Moe Jette
UNICOS 8.0 Installation and Disk Reconfiguration
Accounting Issues
****DISTRIBUTED COMPUTING AND THE NETWORKS****
4:00 ESnet Status Report - Jim Leighton
4:30 ESnet Steering Committee Report - Sandy Merola (LBL)
4:50 Update on Distributed Computing and DCCC/DICCE - Barry Howard
DCCC Petaflop computer
5:15 Adjourn - Dinner at the Hayward Fishery
(EXERSUG members & NERSC management
Wednesday, June 14, 1995
PHYSICS TALK BY HOST INSTITUTION
8:30 John Bell, Center for Computational Sciences and Engineering
- Local Adaptive Mesh Refinement for Computational Fluid
Dynamics
9:00 Bruce Cohen, Numerical Tokamak Project Leader, LLNL
- HPCC Numerical Tokamak Project
9:30 BREAK
9:40 Open Discussion
Invitation to Discussion - Bill McCurdy
Possible Major (and some Minor) Topics Include:
A. ExERSUG/ERSUG involvement in NERSC Move proposal review.
- should we be involved?
- is it possible ?
- who and how ?
B. The case for SMPs
Ideas about increasing the number of compute cycles/dollar
spent at NERSC.
C. The role of the current capability platforms
D. Dial-up Services for NERSC Customers (SLIP, XRemote, PPP)
E. Batch Throughput Issues:
long queue delays
poor throughput
high price/performance relative to interactive computing
F. How to improve participation from the ERSUG community
I. Next ERSUG meeting