Read Format Presentations text version

Defense Programs

Predictive Science Academic Alliance Program II

Thuc Hoang May 10, 2011

PSAAP II Pre-Proposal Conference

May 10-11, 2011

1

Some Historical Background of ASC

Defense Programs

FY94

The President and Congress established the DOE Stockpile Stewardship Program to meet the challenge of maintaining our nuclear deterrent under a comprehensive test ban.

FY96

The ASCI program was established and funded to "promptly shift from nuclear test-based methods to computational-based methods."

FY04

ASCI transitioned to ASC with a vision to "Predict with confidence the behavior of nuclear weapons through comprehensive science-based simulations."

PSAAP II Pre-Proposal Conference May 10-11, 2011

2

Our Business: Using simulation and computing to maintain the US nuclear deterrent

Defense Programs

Central Problem: Replacement of underground testing with a rigorous scientific methodology with which to assess and maintain our confidence in our nuclear stockpile. Time Urgencies: Supporting national policy with respect to the maintenance of our nuclear stockpile requires that we be able to certify annually to the Secretaries of the Departments of Energy and Defense that the stockpile is safe, reliable and secure.

ASC Solution: A National Program planned and coordinated across the three Defense Program Laboratories with partnerships with other federal agencies, academia and industry.

3

ASC Computing Resources Enable Improved Weapons Physics Capabilities

Defense Programs

Stockpile

Alternate materials 2-D SFIs

Energy Balance 3D UGT configurations

Urban consequences Low fidelity, single physics High fidelity, multi-physics Improvised weapon technology Rich physics, 3D

Advanced surety features 3D SFI assessment

Security

Real time response Render safe operations

Basic physics, 1D, 2D

Science

Hydrodynamic instabilities

Laser-plasma interaction Ejecta production Predictive material models

ICF asymmetries

1997

ASC Systems 1TF Red

2003

20TF Q

2005

100TF Purple

2007

500TF BG/L

2009

1PF RoadRunner

2012

20PF Sequoia

2015

100PF

2018

1EF

4

SSP requirements are driving greater understanding, requiring ~100 Exaflops to establish predictive 3D UQ

07 08 09 10

Energy Balance

Defense Programs

23 24 25

11

12

Initial Conditions For Boost Hydro

13

14

15

16

17

18

19

20

21

22

SSP Drivers:

Initial Si Radiation Conditions Damage For Boost Nuclear

Boost Fracture

Secondary Performance

Resolve remaining simulation unknowns

Computer procurements: ASC Cielo

1PF

Establish predictive UQ and key capabilities

ASC Sequoia

10-20PF 150PF 1-2EF 10EF 50-100EF

Science simulations: Design simulations:

Strength data through DD 3D UQ study

Turbulence DNS Re = 300K 3D Science based models

3D IDC atomistic 3D High resolution Capability simulation

QMD EOS data table 3D Predictive UQ study

5

Exascale Computing Is Essential for Nuclear Security Within the Next Decade

Defense Programs

· Stockpile Stewardship critically dependent on predictive capability

As the stockpile is reduced and optimized, single points of failure become critical to understand and resolve quickly Full 3-D parametric simulations (predictive Uncertainty Quantification) requires computing capacity beyond what is feasible on current machines Some improved surety options cannot be certified and implemented without exascale

· US government investment required to support breakthroughs in power,

memory, and SW technologies

Power constraints due to current technologies are not feasible Current production applications need to be re-written/re-packaged for future architectures Current software architecture will not scale New architecture-aware algorithms will be required Pipeline for future workforce at NNSA national labs is not as full as it has been in the past

ASC Exascale Planning- https://asc.llnl.gov/exascale/

PSAAP II Pre-Proposal Conference May 10-11, 2011

6

Next ASC Alliance Program

Defense Programs

·

Focus on a multi-scale, unclassified application of NNSA interest: Multi-disciplinary Simulation Centers (MSCs) Single-Discipline Centers (SDCs)

· ·

Demonstrate a verified, validated predictive simulation capability with uncertainty quantification Establish Exascale computing as an important area of academic research

BLUE = New characteristics in PSAAP II vs. PSAAP I

PSAAP II Pre-Proposal Conference

May 10-11, 2011

7

PSAAP II Boundaries

Defense Programs

·

Budget Structure:

­ $4M/year for MSCs for 5 years ­ $2M/year for SDCs for 5 years ­ Possible 3-year extension upon further demonstration of success

·

Proposers' Qualification:

­ ­ ­ ­ ­ US Ph.D. granting institutions only Minimum of 10% cost matching by awardees Can team up with other universities with one (& only) serving as primary No limit on pre-proposal submissions per university campus PSAAP I Centers are required to propose new and different application problem

PSAAP II Pre-Proposal Conference

May 10-11, 2011

8

PSAAP II Procurement

·

Timeline:

Defense Programs

· · · · · ·

RFI issuance ­ end of June 2011 RFP issuance ­ early CY 2012 Proposal Evaluation & Site Reviews ­ summer 2012 Selection of Awardees ­ end of CY 2012 New PSAAP II Centers ­ mid April 2013

ASC will be seeking interested proposers' inputs via upcoming RFI

Inputs/comments will help in final scoping of the new program Feedback on strengths and weaknesses of pre-proposals will be provided

Track the RFI & RFP on grants.gov: # DE-FOA-0000528

PSAAP II Pre-Proposal Conference

May 10-11, 2011

9

Rules of Engagement/Interaction

Defense Programs

·

Prior to Release of RFI:

­ Proposers are allowed to contact NNSA ASC lab experts to ask questions ­ Word of Caution: View the lab experts' responses or comments as their own, not representative of HQ' or the overall program's or "what to propose" ­ Proposers will be evaluated entirely on their own team's merits, not any labs' nor non-academic collaborators' efforts

·

There should be no discussion wrt this procurement between ASC and any PSAAPII proposal teams between the release of the RFI and the announcement of the results of the RFI.

­ Tentatively scheduled for end of June ­ end of Sept. 2011

·

There should be no discussion between wrt this procurement ASC and any PSAAPII proposal teams between the release of the RFP and the announcement of the final decisions of the RFP.

­ Tentatively scheduled for early CY 2012 ­ end of CY 2012

PSAAP II Pre-Proposal Conference

May 10-11, 2011

10

ASC Teams @ this Conference

Defense Programs

- Alliance Strategy Team (AST): Blue - Verification and Validation (V&V): Red - Multiphysics Applications (Apps): Green - Physics and Engineering Modeling (PEM): Yellow - Computer Science (CS): Orange - ASC Managers: Purple

PSAAP II Pre-Proposal Conference

May 10-11, 2011

11

Defense Programs

[email protected] 202-586-7050

PSAAP II: sandia.gov/ascppc

ASC Program: nnsa.energy.gov/asc

PSAAP II Pre-Proposal Conference

May 10-11, 2011

12

Information

Format Presentations

12 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

97905