HPCMP Objectives Department of Defense High Performance Computing Modernization Program Andrew Mark Program Manager, Applications Software

Program Objectives  Establish a world class capability, within the DoD, to apply high performance computation to solve DoD problems  Ensure military advantage and warfighting superiority on the 21st century battlefield through the use of high performance information technologies  Strengthen national prominence and preeminence by advancing critical technologies and expertise in high performance computing  Support the HPC needs of the defense science & technology (S&T) and test and evaluation (T&E) communities  Establish a world class capability, within the DoD, to apply high performance computation to solve DoD problems  Ensure military advantage and warfighting superiority on the 21st century battlefield through the use of high performance information technologies  Strengthen national prominence and preeminence by advancing critical technologies and expertise in high performance computing  Support the HPC needs of the defense science & technology (S&T) and test and evaluation (T&E) communities

HPCMP Goals  Provide the best commercially available high-end HPC capability  Provide high performance computing environments and software applications that enables critical DoD research, development and test problems to be solved.  Educate and train DoD’s scientists and engineers to effectively use advanced computational environments  Link users and computers sites via high-capacity networks, facilitating user access and distributed computing environments  Promote collaborative relationships among the DoD HPC community, the National HPC community and Minority Serving Institutions (MSIs) in network, computer, and computational science  Provide the best commercially available high-end HPC capability  Provide high performance computing environments and software applications that enables critical DoD research, development and test problems to be solved.  Educate and train DoD’s scientists and engineers to effectively use advanced computational environments  Link users and computers sites via high-capacity networks, facilitating user access and distributed computing environments  Promote collaborative relationships among the DoD HPC community, the National HPC community and Minority Serving Institutions (MSIs) in network, computer, and computational science

A Quick History  Program initiation — 1992–93  HPC Modernization Plan  Initial program structure established  Initial HPC capabilities provided  Program formalization — 1994–95  Program office established  DoD oversight process implemented  Program structure and customer base expanded  Major acquisitions — 1995–96  Four major shared resource centers (MSRCs)  Defense Research and Engineering Network (DREN)  Program initiation — 1992–93  HPC Modernization Plan  Initial program structure established  Initial HPC capabilities provided  Program formalization — 1994–95  Program office established  DoD oversight process implemented  Program structure and customer base expanded  Major acquisitions — 1995–96  Four major shared resource centers (MSRCs)  Defense Research and Engineering Network (DREN)  Program upgrades and operations — 1997–2000  Continuous upgrades at MSRCs  Selection of new distributed centers (DCs)  DoD Challenge Projects established  A new round of acquisitions — 2001–03  New DREN contract let with WorldCom  New support contracts at MSRCs  Annual technology insertion process at MSRCs  Program upgrades and operations — 2004–09

Current User Base and Requirements  561 projects and 4,572 users at approximately 179 sites  Requirements categorized in 10 Computational Technology Areas (CTA)  FY 2005 non-real-time requirements of 260 teraFLOPS-years  561 projects and 4,572 users at approximately 179 sites  Requirements categorized in 10 Computational Technology Areas (CTA)  FY 2005 non-real-time requirements of 260 teraFLOPS-years CSM – 507 Users CFD – 1,135 Users CCM – 235 Users CEA – 304 Users CWO – 231 Users SIP – 435 Users FMS – 889 Users EQM – 170 Users IMT – 568 Users 60 users are self characterized as “other” CEN – 38 Users

Computational Technology Areas (CTAs) Data Source: HPCMP Requirements Analysis, Forces Modeling and Simulation (FMS)/C4I Michael Macedonia, PEO STRI 889 1,412 Environmental Quality Modeling and Simulation (EQM) Jeff Holland, ERDC Computational Electronics and Nanoelectronics (CEN) Barry Perlman, CERDEC Integrated Modeling and Test Environments Jere Matty, AEDC 568 1,275 Other

HPC Modernization Program User Base

HPCMP Software Test Process  Monitor Implementation  Provide Workshops/ Training/Templates/Examples  Review/Approve Test Plans  Implementation/Execution  Feedback to Site  Independent Evaluation & Reports  Conducts Test Plan Workshops  Reviews Test Plans  Witnesses Tests, Reviews Test Results & Reports, Provides Feedback to HPCMPO & DOT&E Develops & Executes:  System Acceptance Test Plans  Full Operational Capability Test Plans  System & FOC Assessment Tests  Provides Test Reports SITE JITC HPCMPO

HPC Software Applications Institutes and Focused Portfolios  5 HPC Software (Applications) Institutes  HPCMP chartered  Service managed  6 year duration  Ends with Transition to Local Support  $2M annual funding for:  3-12 computational and computer scientists  Support development of new and existing codes  Adjust local business practice to use science- based models & simulation  Integrated with PET  5 HPC Software (Applications) Institutes  HPCMP chartered  Service managed  6 year duration  Ends with Transition to Local Support  $2M annual funding for:  3-12 computational and computer scientists  Support development of new and existing codes  Adjust local business practice to use science- based models & simulation  Integrated with PET Service Management PET ON-SITES HPC Software Applications Institutes $8-12M Portfolio MGR HPC-SA MGR Project Team Computational Projects Service Management Service Management Service Management Service Management

HPC Software Portfolios   Provides core software applications, models, and simulations which exploit HPC capabilities in areas of most concern to DoD   Multi-purpose codes perform on a range of HPC platforms, with focus on accuracy, efficiency, reusability, and scalability   Inter-Service, multi-disciplinary teams include algorithm and code developers, applications specialists, computer scientists and end users who are responsible for development, dissemination, technology transfer and follow-on support   Collaboration with PET academics and major shared resource centers on innovative software tools, training, and other elements common to HPC programming environments   Provides core software applications, models, and simulations which exploit HPC capabilities in areas of most concern to DoD   Multi-purpose codes perform on a range of HPC platforms, with focus on accuracy, efficiency, reusability, and scalability   Inter-Service, multi-disciplinary teams include algorithm and code developers, applications specialists, computer scientists and end users who are responsible for development, dissemination, technology transfer and follow-on support   Collaboration with PET academics and major shared resource centers on innovative software tools, training, and other elements common to HPC programming environments

Institute/Portfolios/CTA Software Goals  Enhance existing service/agency high priority projects with software development support by institutionalizing a computational culture to ensure that the resulting software applications effectively and efficiently utilize the most current HPC technology  Utilize science and technology resources to develop, apply and transition DoD HPC software applications in order to:  Lower cost and accelerate weapons systems design, development, testing, and acquisition  Improve readiness and deployment through the use of simulation-based mission rehearsal, training, situational awareness and environmental analysis  Lower the cost of software maintenance  Provide for a greater usability and robustness of software across a wide user base  Build HPC experience in defense laboratories and test centers for defense applications among DoD, industrial, and academic scientists and engineers  Leverage other HPC efforts across DoD, government, industry, and academia  Enhance existing service/agency high priority projects with software development support by institutionalizing a computational culture to ensure that the resulting software applications effectively and efficiently utilize the most current HPC technology  Utilize science and technology resources to develop, apply and transition DoD HPC software applications in order to:  Lower cost and accelerate weapons systems design, development, testing, and acquisition  Improve readiness and deployment through the use of simulation-based mission rehearsal, training, situational awareness and environmental analysis  Lower the cost of software maintenance  Provide for a greater usability and robustness of software across a wide user base  Build HPC experience in defense laboratories and test centers for defense applications among DoD, industrial, and academic scientists and engineers  Leverage other HPC efforts across DoD, government, industry, and academia

High Performance Computing Software Applications Institute for Space Situation Awareness ( Air Force Research Laboratory)

Institute for Maneuverability and Terrain Physics Simulation (Engineer Research and Development Center) Why this institute? This institute will:  Foster a culture in which high-fidelity simulation is used to support DoD maneuver dominance by  Detecting landmines, improvised explosive devices (IED), and unexploded ordnance (UXO),  Analyzing maneuver and trafficability, and  Remotely sensing denied areas and using unattended ground sensor networks. Why this institute? This institute will:  Foster a culture in which high-fidelity simulation is used to support DoD maneuver dominance by  Detecting landmines, improvised explosive devices (IED), and unexploded ordnance (UXO),  Analyzing maneuver and trafficability, and  Remotely sensing denied areas and using unattended ground sensor networks.  Develop a virtual testing facility based on state-of-the-ground analyses  Develop inverse models to predict ground state of denied areas through remote sensing  What are the biggest challenges?  Integration of ground physics, sensor physics, and sensor response models to facilitate virtual testing of sensor performance  Inverse modeling to estimate the ground state of denied areas using remote sensing and coupled, multi-physics models

Biotechnology HSAI for Force Health Protection (U.S. Army Medical Research and Materiel Command (MRMC)) Institute Objectives  Specific objectives include:  Development and application of HPC software (of military relevance) for the analysis of genomic and proteomic data  Expansion of a cross-trained workforce with modern computational skills and improved biomedical knowledge  Leveraging the powerful synergistic relationship between laboratory experimentation and HPC simulation  The Institute will have a unique mission within the DoD Institute Objectives  Specific objectives include:  Development and application of HPC software (of military relevance) for the analysis of genomic and proteomic data  Expansion of a cross-trained workforce with modern computational skills and improved biomedical knowledge  Leveraging the powerful synergistic relationship between laboratory experimentation and HPC simulation  The Institute will have a unique mission within the DoD The goal is to serve as a tri-Service, inter-disciplinary resource to accelerate R&D of countermeasures for Force Health Protection

Battlespace Environments Institute (Naval Research Laboratory) Institute Objectives  To develop a DoD-wide whole-earth environment which interoperates with that from other agencies:  Migrate core models to ESMF  Navy (e.g., NCOM, HYCOM, SWAN, COAMPS™)  Air Force (Kinematic Solar Wind and GAIM )  Army (e.g., ADCIRC, WASH123, multi-utility ADH)  WRF and CICE are founding ESMF components  Integrate technology (e.g., movable nests, aerosol analysis, tropical cyclone initialization etc.) from COAMPS™ into WRF.  Development of tools and applications (e.g., extend ESMF to support unstructured grids and nesting)  Coupled applications: Air/ocean, air/ocean/ice, air/ocean/groundwater, air/space-weather  Thorough testing, prototyping, documentation of all components Institute Objectives  To develop a DoD-wide whole-earth environment which interoperates with that from other agencies:  Migrate core models to ESMF  Navy (e.g., NCOM, HYCOM, SWAN, COAMPS™)  Air Force (Kinematic Solar Wind and GAIM )  Army (e.g., ADCIRC, WASH123, multi-utility ADH)  WRF and CICE are founding ESMF components  Integrate technology (e.g., movable nests, aerosol analysis, tropical cyclone initialization etc.) from COAMPS™ into WRF.  Development of tools and applications (e.g., extend ESMF to support unstructured grids and nesting)  Coupled applications: Air/ocean, air/ocean/ice, air/ocean/groundwater, air/space-weather  Thorough testing, prototyping, documentation of all components

Institute for HPC Applications to Air Armament (IHAAA) (Air Force SEEK EAGLE Office)  Why IHAAA? Revolutionize Arming the Warfighter  Optimize/expand HPC use in Air Armament  Quick turn user operational requests  Reduce lifecycle risks, costs, schedules  T&E Centric Proposal  Customer focus –  Warfighter, Acquisition  Capability not just code – we are end users  This is about process not just technology  Build the research warfighter bridge  Joint processes – Shared people, tools, knowledge, and usable models  Seamless HPC model transition between acquisition phases  Proven capability: high probability of success and sustainment  Army, Navy, Air Force, and Joint Strike Fighter PO Endorsed  Why IHAAA? Revolutionize Arming the Warfighter  Optimize/expand HPC use in Air Armament  Quick turn user operational requests  Reduce lifecycle risks, costs, schedules  T&E Centric Proposal  Customer focus –  Warfighter, Acquisition  Capability not just code – we are end users  This is about process not just technology  Build the research warfighter bridge  Joint processes – Shared people, tools, knowledge, and usable models  Seamless HPC model transition between acquisition phases  Proven capability: high probability of success and sustainment  Army, Navy, Air Force, and Joint Strike Fighter PO Endorsed

DoD R&D Laboratories and T&E Centers DoD R&D Laboratories and T&E Centers Commercial Commercial Off-The-Shelf Software Commercial Commercial Off-The-Shelf Software ComputerSystems l Vector l Scalable l Clustered ComputerSystems l Vector l Scalable l Clustered Storage Storage l Distributed File Systems Systems l Robotics Archival Systems Systems Storage Storage l Distributed File Systems Systems l Robotics Archival Systems Systems Scientific ScientificVisualization Visualization Internal Networking NetworkingNetworking High Performance Computing Centers Centers DoD HPC DoD HPCModernization Program Program DoD HPC DoD HPCModernization Program Program Defense Research & EngineeringNetworkDefense EngineeringNetwork Software Applications Support Software Integrated Program Strategy S&T & T&E ApplicationsS&T Applications DevelopmentSupportDevelopmentSupport Program Components Programming Programming Environment and Training Programming Programming Environment and Training Computer Science Services Services