Applied Mathematics Research Program SIAM Computational Science & Engineering March 5, 2009 Sandy Landsberg, Program Manager Office of Advanced Scientific.

Applied Mathematics Research Program

SIAM Computational Science & EngineeringMarch 5, 2009

Sandy Landsberg, Program ManagerOffice of Advanced Scientific Computing Research

Office of ScienceDepartment of Energy

U.S. Department of Energy’sOffice of Science

207/11/2008 SIAM Annual Meeting 2008 DLB-2

DOE Mission Elements

• Energy • Nuclear power, clean coal, fusion reactors,

enhanced oil recovery, bio-fuels, reliability & security of electric power grid, new engine designs

• Environment • Carbon sequestration, nuclear waste storage,

environmental cleanup, climate research• National Security

• Nuclear stockpile stewardship

307/11/2008 Applied Math at the DOE David L. Brown

What kinds of questions can CS&E help answer for DOE?

• Can we predict the operating characteristics of a clean coal power plant?

Answering these and other important questions involves study of increasingly complex physical and engineered systemsAnswering these and other important questions involves study of increasingly complex physical and engineered systems

• How stable is the plasma containment in a Tokamak?

• How quickly is climate change occurring and what are the uncertainties in the predicted time scales?

• How quickly can the US recover if part of the power grid became inoperable?

• How can new materials be designed with a specified desirable set of properties?

4

FY 2008Approp.

FY 2009Request

Applied Mathematics 36,885 43,164

Computer Science 27,226 34,618

Computational Partnerships 53,767 52,064

Next Gen. Networking for Science 12,017 17,221

High Performance Production Computing 54,200 54,790

Leadership Computing Facilities 110,158 115,000

High Performance Network Facilities & Testbeds 23,936 25,000

Research and Evaluation Prototypes 23,585 17,000

Subtotal, ASCR 341,774 358,857

All other 9,399 9,963

Total, ASCR 351,173 368,820

Office of Advanced Scientific Computing Research (ASCR)

Mission:Discover, develop, and deploy the computational and networking tools that enable researchers in the scientific disciplines to analyze, model, simulate, and predict complex phenomena important to the Department of Energy.

5

It’s not all about hardware …

CS&E advances are essential for the exponential performance increases that drive scientific discovery through computation

0

1

2

3

4

5

6

7

8

9

10

1980 1990 2000 2010

Calendar Year

Lo

g E

ffe

cti

ve

Gig

aF

LO

PS

High Order

Autocode

ARK integratorcomplex chem Higher

orderAMR

NERSCRS/6000

NERSCSP3

Cray 2

AMR

Low Mach

Combustion modelingModeling advances

Algorithm advances

Hardware advances

Excerpt from: A Science-Based Case For Large Scale Simulation Volume 2http://www.pnl.gov/scales/docs/SCaLeS_Report_Vol2.pdf

6

Applied Mathematics Research Program

• Currently supported research activities (~ 100 projects):

• Advanced linear algebra • Discretization and meshing • Multiscale, multiphysics systems• UQ and error analysis• Optimization• Other research• Fellowships & workshops

• Outreach to mathematics, computer science and computational science community

von Neumann 1903-1957

Research on mathematical models, methods and algorithms to enable scientists to accurately understand complex physical, chemical, biological and engineered systems.

7

Applied Mathematics Research New Research Thrusts

• Multiscale Mathematics & Optimization for Complex Systems• http://www.sc.doe.gov/ascr/Research/AM/ComplexSystemsWorkshopReport.pdf• http://www.sc.doe.gov/ascr/Research/AM/MultiscaleMathWorkshop1.pdf• http://www.sc.doe.gov/ascr/Research/AM/MultiscaleMathWorkshop2.pdf• http://www.sc.doe.gov/ascr/Research/AM/MultiscaleMathWorkshop3.pdf

• Mathematics for Analysis of Petascale Data• http://www.science.doe.gov/ascr/ProgramDocuments/Docs/PetascaleDataWorksh

opReport.pdf

• Joint Mathematics / Computer Science Institute(s)• http://www.csm.ornl.gov/workshops/institutes/

• Mathematics of Cybersecurity• http://www.science.doe.gov/ascr/ProgramDocuments/Docs/

CyberSecurityScienceDec2008.pdf• http://csmr.ca.sandia.gov/~tgkolda/pubs/bibtgkfiles/SAND2009-0805-eco%20Math

%20for%20Cyber.pdf

• Unsolicited proposals: • http://www.science.doe.gov/grants/FAPN09-01.html

8

CS&E Community Involvement

• Learn more about ASCR programs

• Be a reviewer• Panel reviews• “Postal” reviews

• Participate in DOE workshops• ASCR-sponsored• Basic Energy Sciences (BES), Biological and

Environmental Research (BER), Electricity Delivery and Energy Reliability (OE), Energy Efficiency and Renewable Energy (EERE), Nuclear Energy (NE)

• Identify future research needs• Generate “strong” community interest

• ACCOMPLISHMENTS!

9

New “Breakthroughs” report

• Each of Mathematics, Computer Science, and Applications will have a “breakthroughs” report to support FY11 DOE budget request

• Each report will look back about eight years – the SciDAC era

• What are the ten most important mathematical breakthroughs to computational science?

• Not necessarily scaling issues, could be “more science per flop/byte stored/byte transmitted”

• Please send your suggestions to

[email protected]

10

Questions?

Sandy [email protected]

301-903-8507

Steve [email protected]

301-903-5710

Karen [email protected]

301-903-5384

Bill [email protected]

301-903-8268

http://www.science.doe.gov/ascr/Research/AppliedMath.html