ASC Contributions to the Stockpile Stewardship Program (SSP)
In FY 2015, ASC worked to improve its unique integrated design codes with a focus on sufficient resolution, dimensionality, and scientific detail necessary to address the increasingly difficult analyses needed for stockpile stewardship. Modern ASC simulation codes are now used for performance assessment of all enduring stockpile systems. The use of ASC codes along with common-model methods and updated material properties resulted in significant changes to performance assessments. Reevaluation of one Limited-life Component (LLC) using ASC codes resulted in an extension of its serviceable life. This change will help to avoid schedule delays and cost increases in the replacement of that LLC. Potential future Significant-Finding-Investigation (SFI) issues were identified in a proposed W88 Alt component design through the application of uncertainty analysis enabled by ASC tools. The ASC Cielo system was used in support of critical B61-12 LEP deliverables for engineering design and assessment. To address issues associated with new computer architectures, the three ASC laboratories collaborated on an analysis using proxy applications to assess performance gains and programming abstractions. The Sierra supercomputer and the Commodity Technology System (CTS-1) contracts were awarded. Two Centers of Excellence have been created to assist with ASC application porting to new advanced technology platforms. Software packages are being developed to address key technical challenges, in particular, memory hierarchies, programming models, power, and resilience. A contract was awarded for a new modular unclassified computing facility at LLNL, the power and cooling were reconfigured to support the new Trinity supercomputer housed at LANL’s Strategic Computing Complex, and networks were upgraded at LANL and between SNL and LANL.
In FY 2014, ASC continued delivering science-based simulation tools for annual assessments and next-generation LEPs, focusing on improved physics, fidelity, and calculations in support of DSW and the National Code Strategy. The IC subprogram continued to add capabilities to their physics and engineering codes. A new 3D adaptive mesh refinement (AMR) capability was successfully applied to model the behavior of additively manufactured parts under extreme conditions. Within IC, ASC’s collaborations with academia via the Predictive Science Academic Alliance Program (PSAAP II) continued with six newly selected Centers. Some recent developments from the ASC PEM subprogram in FY14 include 1) reactive molecular dynamics simulations on an unprecedented scale (length ~1 micron, time ~1 ns), confirming the hypothesis that voids play a critical role in determining the initiation threshold in energetic materials and pave the way to inform grain-scale hydrocode simulations in the future; and 2) predictions of four distinct theories for the equation of state (EOS) of lithium deuteride, providing the capability to create more accurate databases to match these simulations and existing experimental data obtained by the science campaigns in both high-energy-density and focused experiments. Within the CSSE subprogram, the vendor for the first ASC advanced technology system, which is named Trinity and will be deployed at Los Alamos National Laboratory (LANL) in the first quarter of FY16 and capable of 33.6 petaFLOP/s, was selected. The FastForward and DesignForward engagements with industry, in the Advanced Technology Development and Mitigation (ATDM) subprogram, have begun to accelerate industry’s R&D technology roadmap for the program’s future exascale-class computational needs.
In FY 2013, ASC continued delivering science-based simulation tools for annual assessments and next-generation LEPs, focusing on improved physics, fidelity, and calculations in support of DSW and the National Code Strategy. Sequoia, the advanced architecture system, was delivered to Lawrence Livermore National Laboratory (LLNL), and transitioned to the classified environment in the beginning of 2013. This 20-petaFLOP/s (PF) supercomputer will be focused on strengthening the foundations of predictive simulation through running very large suites of complex simulations called uncertainty quantification studies. In addition, it will be used for weapons science calculations necessary to build more accurate physical models. Procurement of the next-generation tri-lab Linux capacity clusters (TLCC2) and the associated common user environment milestones have continued to proceed as planned. However, the acquisition of the commodity technology systems (CTS) that will replace the TLCC2 capabilities in FY14 has been pushed back a year at all three of the NNSA laboratories, with an FY15 start. Roadrunner, the first 1-PF advanced architecture system at Los Alamos National Laboratory (LANL), was retired. A request for proposals was released for the acquisition of an advanced technology system (ATS) to be deployed at LANL in the first quarter of FY16, named Trinity, which will be in the 40–90 petaFLOP/s range. Responses to ASC’s RFPs were evaluated for the next round of Predictive Science Academic Alliance Program (PSAAP II) engagements, selection activities were conducted, and recipients of these awards were selected.
In FY 2012, ASC released a Request for Proposal and reviewed proposals submitted for the PSAAP II program. Cielo, the new capability machine in operations at LANL, entered the General Availability phase in September 2012. All 96 racks of Sequoia were delivered in May 2012 and were integrated on the unclassified network by the end of the fourth quarter. Integration and operation of TLCC2 systems deployed across the tri labs were implemented by end of the fiscal year. FastForward investments, co-funded with DOE Office of Science Office of Advanced Scientific Computing Research to accelerate extreme-scale computing R&D technologies, were awarded to AMD, IBM, Intel, Nvidia, and WhamCloud. On the Physics and Engineering Models (PEM) front, a Level 1 milestone to advance capabilities for annual assessments and resolution of SFIs associated with early-phase primary implementation progressed as planned. Verification and validation assessment of improvements in primary performance codes for boost continued. The Verification and Validation (V&V) subprogram also worked on the strategy of common modeling to validate improvements in support of the National Boost Initiative. The Integrated Codes (IC) program element concentrated on preparations for the next predictive capability framework (PCF) peg post and on a Level 1 milestone for simulating late-time primary implosion and initial explosion, as well as on scalability enhancements targeting future computing platforms.
In FY 2011, ASC continued delivering science-based simulation tools for annual assessments and next-generation life extension programs (LEPs), focusing on improved physics, fidelity, and calculations in support of Directed Stockpile Work (DSW) and the National Code Strategy. The methodology for predictive capability assessment was demonstrated in FY11 for a limited set of simulations common to both physics laboratories. ASC assessed the ability to simulate full system performance near thresholds where data are sparse. Cielo, the capability HPC system, replaced Purple and began operation. Cielo was upgraded to 1.37 petaFLOP/s, the system was accepted, and the second round of the Capability Computing Campaign (CCC) was initiated. Sequoia, the advanced architecture system, received a “go” from the program’s “go/no go” review. Contract negotiations and delivery schedule have been finalized. As of September 30, 2011, facility site preparation work was well underway, and design and procurement of the parallel file system was complete. Building on the FY10 Scalable Applications Project (SAP) milestone, the SAP effort has extended the knowledge base, documentation, and training to enable tri-lab code teams to use Sequoia. Successful procurement of the next-generation tri-lab Linux capacity clusters (TLCC2) and the associated common user environment milestones were executed across the three NNSA laboratories. Initial investments were made towards the proposed joint DOE Office of Science (SC) and NNSA exascale initiative. Proposal responses to ASC’s Request for Information were evaluated in preparation for the next round of Predictive Science Academic Alliance Program (PSAAP) engagements.
In FY 2010, ASC continued to deliver science-based simulation tools to support annual assessments and the next generation of LEPs. Code suite physics and optimization were completed in support of the NTNF program and High Energy Density Physics experimental program. The Energy Balance Level 1 milestone was completed using ASC codes and the Predictive Capability Assessment Project was begun under the V&V program element. An initial assessment of new capabilities in a primary burn code was performed. ASC also provided tools for both experiment and diagnostic design to support the indirect-drive ignition experiments on the NIF. In addition, ASC continued to provide national leadership in HPC and deploy capability and capacity platforms in support of DP Campaigns. Roadrunner, the advanced architecture petaFLOP hybrid HPC was formally transitioned to production computing for weapons applications.
In FY 2009, ASC released improved codes to support stockpile stewardship and other nuclear security missions, including secure transportation, NSE infrastructure, and nuclear forensics—specifically, a suite of physics-based models and high-fidelity databases were developed and implemented to support National Technical Nuclear Forensics (NTNF) activities.
In FY 2008, ASC delivered the codes for experiment and diagnostic design to support the CD-4 approval on the National Ignition Facility (NIF). An advanced architecture platform capable of sustaining a 1-petaFLOPS benchmark, named Roadrunner, was sited at Los Alamos National Laboratory (LANL). SNL and LANL established the collaborative Alliance for Computing at Extreme Scale (ACES) for the purpose of providing a user facility for production capability computing to the Complex. Plans were made for the Cielo capability computing platform, the first platform to be hosted through ACES, to be procured and sited at LANL.
In FY 2007, ASC supported the completion of the W76-1 and W88 warhead certification, using quantified design margins and uncertainties; ASC also provided two robust 100-teraFLOPS-platform production environments by IBM and CRAY, supporting DSW and Campaign simulation requirements, respectively. One of the original ASCI program Level 1 milestones was completed when the ASC Purple system was formally declared “generally available.” This was augmented by the 360-teraFLOPS ASC BlueGene/L system, which provided additional capability for the Science Campaigns. The ASC funded partnerships with Sandia National Laboratories (SNL)/Cray and Lawrence Livermore National Laboratory (LLNL)/IBM have transformed the supercomputer industry. By mid-2007, there were at least 34 “Blue Gene Solution” systems on the Top 500 list and 38 Cray sales based on the SNL Red Storm architecture.
In FY 2006, ASC delivered the capability to perform nuclear performance simulations and engineering simulations related to the W76/W80 LEPs to assess performance over relevant operational ranges, with assessments of uncertainty levels for selected sets of simulations. The deliverables of this milestone were demonstrated through two-dimensional (2D) and 3D physics and engineering simulations. The engineering simulations analyzed system behavior in abnormal thermal environments and mechanical response of systems to hostile blasts. Additionally, confidence measures and methods for UQ were developed to support weapons certification and QMU Level 1 milestones.
In FY 2005, ASC identified and documented SSP requirements to move beyond a 100 teraFLOPS computing platform to a petaFLOPS-class system; ASC delivered a metallurgical structural model for aging to support pit-lifetime estimations, including spiked-plutonium alloy. In addition, ASC provided the necessary simulation codes to support test readiness as part of NNSA’s national priorities.
In FY 2004, ASC provided simulation codes with focused model validation to support the annual certification of the stockpile and to assess manufacturing options. ASC supported the life-extension refurbishments of the W76 and W80, in addition to the W88 pit certification. In addition, ASC provided the simulation capabilities to design various nonnuclear experiments and diagnostics.
In FY 2003, ASCI delivered a nuclear safety simulation of a complex, abnormal, explosive initiation scenario; ASCI demonstrated the capability of computing electrical responses of a weapons system in a hostile (nuclear) environment; and ASCI delivered an operational 20-teraFLOPS platform on the ASCI Q machine, which has been retired from service.
In FY 2002, ASCI demonstrated 3D system simulation of a full-system (primary and secondary) thermonuclear weapon explosion, and ASCI completed the 3D analysis for an STS abnormal-environment crash-and-burn accident involving a nuclear weapon.
In FY 2001, ASCI successfully demonstrated simulation of a 3D nuclear weapon secondary explosion; ASCI delivered a fully functional Problem Solving Environment for ASCI White; ASCI demonstrated high-bandwidth distance computing between the three national laboratories; and ASCI demonstrated the initial validation methodology for early primary behavior. Lastly, ASCI completed the 3D analysis for a stockpile-to-target sequence for normal environments.
In FY 2000, ASCI successfully demonstrated the first-ever 3D simulation of a nuclear weapon primary explosion and the visualization capability to analyze the results; ASCI successfully demonstrated the first-ever 3D hostile-environment simulation; and ASCI accepted delivery of ASCI White, a 12.3-teraFLOPS supercomputer, which has since been retired from service.
In FY 1998, ASCI Blue Pacific and ASCI Blue Mountain were delivered. These platforms were the first 3-teraFLOPS systems in the world and have both since been decommissioned.
In FY 1996, ASCI Red was delivered. Red, the world’s first teraFLOPS supercomputer, was upgraded to more than 3 teraFLOPS in FY1999 and was retired from service in September 2005.