the nersc center --from a doe program manager’s perspective--
Post on 02-Jan-2016
19 Views
Preview:
DESCRIPTION
TRANSCRIPT
1
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
The NERSC Center
--From A DOE Program Manager’s Perspective--
A Presentation to the
NERSC Users Group
Walter M. PolanskyActing Program Manager for NERSC
LBNL
February 22, 2001
2
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
Presentation Topics
• MICS
• NERSC Center
• Budgets
• The Greenbook
• Challenges/Opportunities
3
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
Need for Information Technology R&D
• Vendors cannot meet requirements for high-performance computing because DOE science programs
require computing capabilities 10 to 100 times greater than typically available in commercial systems.
move hundreds to thousands of gigabytes of data between only a few locations.
4
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
• Many scientific questions important to DOE involve the integration of many scientific disciplines with characteristic lengths and times that range over many orders of magnitude (e.g. combustion).
• Computer simulation is the only practical way to study some systems (e.g. fusion plasma, climate, genome).
• Computer simulation can complement data and provide insight into the behavior of systems.
• High performance networks are critical to harness distributed resources for the analysis of experimental data (millions of gigabytes/year) and to support distributed simulations.
High Performance Computers and Networks -- Importance to DOE Science --
5
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
• Applications, e.g. • Materials Sciences • Chemical Sciences • Combustion Modeling • Accelerator Science • High Energy Physics • Nuclear Physics • Fusion Energy • Global Climate • … • ...
FundamentalResearch
Underlying Technologies
• Hardware
• Software
• Networking
R&D forApplications
ScientificComputing
ASCR
Disciplinary Science and Engineering
External to SC
Energy Sciences Network (ESnet)
Advanced Computing Research Facilities
National Energy Research Scientific Computing Center (NERSC)
Faci
litie
s
Testbeds
• Access to Facilities• Link Researchers
• Applied Mathematics• Computer Science
• Advanced Computing Software Tools
• Scientific Application Pilots
• Networking • Collaboratory Tools • Collaboratory Pilots
BES,BER,FES,HEP,NP
• Enabling Technology CentersGroups of mathematicians, computer scientists, application scientists, and software engineers
MICS is an INTEGRATEDINTEGRATED Program Research --> Users
6
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
MICS Budget History
0
50
100
150
200
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
Fiscal Year
$ (in
Mill
ions
)
Research
Facilities (NERSC+ACRFs+ESnet)
7
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
National Energy Research Scientific Computing Center
-- Goal --
To provide state of the art, high performance computing capability resources and professional user friendly services to a broadly based user community of computational scientists in support of the missions of the Office of Science.
8
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
NERSC-Enabled Research Accomplishments
• Magnetic materials
• Adaptive Mesh Refinement in Complex Geometries
• Accelerator design
• Stellerator optimization
• Predicting Protein Structure
• Microstructural Pattern Formation in Directional Solidification
• Coupled Ocean/Atmosphere System
9
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
Advancing Computational Science of Scale- Producing Real Results
10
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
Budget History- NERSC, ACRF and ESnet at LBNL-
0
10
20
30
40
1997 1998 1999 2000 2001
Fiscal Year
$ (
in M
illio
ns
)
NERSC ACRF ESnet
11
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
• Greenbook objective- Make a compelling case, based on the science, for upgrading NERSC resources (hardware + services).–Quality of science enabled by current NERSC. –Existing and planned research opportunities unmet with present
resources.
• Target Audience- Office of Science program managers.
• Other interested parties- OMB, OSTP, Congress, IG and GAO.
• Attributes- –Cognizance of plans and initiatives (e.g. Flagship-Topical Center
concept, nanotechnology, SciDAC).–Good scientific and technical content; but readable by non-
scientists.–An Executive Summary that (1) provides a concise overview of the
document; (2) and coaxes the reader to continue.
Thoughts on the Greenbook
12
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
The Greenbook- Foundation of a Strategy
Objective Instrument Lead Organization
Establish scientific basis Greenbook NERSC Users Group
Identify SC program needs
--- SC Program Offices
Propose resources needed to meet needs
--- NERSC
Prioritize high-performance computing needs
Budget request; Approved Plans
Office of Science
13
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
• Increase awareness of NERSC-enabled contributions to the Office of Science mission.
• Remain engaged in NERSC planning and operations.
• Share information and experiences with user groups at other computing centers.
• Identify recent major scientific accomplishments and researcher recognition directly influenced by NERSC.
• View NERSC as an SC resource and asset.
Challenges/Opportunities
14
OFFICE OFADVANCED SCIENTIFIC COMPUTING RESEARCH
Personal Areas of Responsibility
•Program Manager for NERSC, Acting•Program Manager for ACRFs, Acting•Program Manager for Scientific Applications Pilot Projects, Acting•Director, Technology Research Division
-Laboratory Technology Research Program-SBIR and STTR Programs
top related