A publication of the Office of Advanced Simulation & Computing, NNSA Defense Programs
NA-ASC-500-07—Issue 3
May 2007
Printable Version(~400KB PDF)
The View from HQ
by Dimitri Kusnezov
I have been spending much of my time these days thinking about science, technology and engineering and the role of the laboratories and how that will be reflected in the Complex of the future. This is on my mind for two reasons: one is my responsibility to produce a science and technology roadmap for Complex 2030—Defense Program’s vision for what we should be doing in 2030—and how we plan to get there. The second reason is to determine what this means for the near-term roles of the laboratories. For some time now I have been concerned that too much of our national security mission is tied to the nuclear weapons monolith, leveraged on the continuation of an enterprise that has not fully transitioned out of the cold-war. Today’s and tomorrow’s threats are far different than what we worried about decades ago, but our mission space has never been redefined to meet our new reality. While the laboratories have recognized some of this, the question here is whether there is a mission space that requires federal land-lording and strategic investment that we are not planning for today. I believe this is the case, and in conjunction with the Complex 2030 Roadmap, I am endeavoring to push this along.
ASC Program Begins Work on National Code Strategy
The ASC Program has embarked on developing a national code strategy designed to help achieve the Program’s goal of delivering a credible and sustainable predictive simulation capability for stockpile stewardship. This code strategy is one of the next major programmatic plans being developed by the tri-lab ASC community and Headquarters. The strategy will be a high-level “design document” for the future code development activities sponsored by ASC, and it complements the larger ASC Roadmap and the Predictive Capability Framework, directed at how we implement the program through our code portfolio and supporting research.
A team of three representatives each from Sandia, Los Alamos, and Lawrence Livermore national laboratories, plus representatives from NNSA Headquarters, completed its first meeting on May 1 and 2, 2007, at Sandia. A wide range of topics was discussed, including an overview of current capabilities, how code and method diversity impact the ability to support stockpile stewardship, and the potential impact of coming changes in computational technologies on the code portfolio.
The near-term goal is to begin developing a vision for ASC’s Integrated Codes program element as it would exist in 2018, and to test that vision against a number of possible future scenarios for the larger Stockpile Stewardship Program and Complex 2030. The team will meet again in June to work on these near-term goals. It expects to complete the full strategy in early 2008.
Report of the Predictive Science Panel Issued: Many Examples of Outstanding and World-Class Science
The 2006–2007 Report of the Predictive Science Panel (PSP) noted “many examples of outstanding and world-class science, ranging from awarding-winning materials science simulations and ASC-informed re-analyses of UGT [underground test] data to very high strain-rate materials strength experiments.” The PSP meets annually at both Lawrence Livermore and Los Alamos national laboratories to evaluate progress toward the goal of a credible predictive capability, answering the question: How do you know it’s right?
ASC Simulation and V&V Showcased at the ICED-07 Conference, Portugal
François Hemez, of X-1-MV at Los Alamos, was one of three keynote speakers at the 4th International Conference on Engineering Dynamics (ICED). The ICED is organized by the Instituto Superior Tecnico, Universidade Tecnica de Lisboa. This year, the conference was held in Carvoeiro, Portugal, from April 16 to 19, 2007. It is an international forum where recent advances in testing, modeling, and validation are discussed for engineering applications. One-hundred engineers and scientists attended, originating, for the most part, from Eastern and Western Europe, and the United States. Hemez’s keynote presentation, entitled “15 Years of Verifying and Validating Simulations at Los Alamos” (LA-UR-07-2213), showcased some of the accomplishments of the ASC Program in terms of simulation capability and Verification and Validation (V&V), with an emphasis on engineering applications. The conference website can be accessed at http://www.dem.ist.utl.pt/~iced2007/.
Red Storm Enters General Availability
Red Storm supercomputer, located at Sandia National Laboratories (Albuquerque, NM) entered the “General Availability” stage on March 26, 2007. This change marked the completion of all initial deployment and testing tasks. The system had been operating in “Limited” availability since September 2005, pending improvements to both the platform and the supporting systems (Red and Black RoSE). The platform upgrades resulted in Red Storm being ranked #2 on the Top 500 list of supercomputers published in November 2006.
During General Availability, all new user requests will be honored if the users have legitimate Nuclear Weapons, Directed Stockpile Work, or Alliance partners’ project needs.
Current user accounts will remain active.
Sandia welcomes this new stage in Red Storm’s activities. The entire Red Storm and Sandia Supercomputing Support teams are looking forward to working with new users to maximize their productivity on the platform.
Critical Decision-0 Approved for ASC Sequoia Supercomputer
The Critical Decision Mission Need Package (CD-0) for the new ASC Sequoia supercomputer was approved and signed on April 23, 2007, by Marty Shoenbauer, acting NNSA Deputy Administrator for Defense Programs. CD-0 approves the NNSA mission need for Sequoia, a new uncertainty quantification (UQ)-focused computer system to be delivered in 2011. The CD-0 package contains a mission need statement, business case, and preliminary project execution plan. It represents the initiation stage of project management activities for Sequoia.
New Multiphase Plutonium Equation of State at Lawrence Livermore Improves Current High Standard and Aids Stockpile Certification
Replacing an equation-of-state (EOS) high standard set in the 1990s for weapons simulation at Lawrence Livermore, a new multiphase plutonium (Pu) EOS will allow weapons designers to elaborate important weapons-relevant physics beyond current capabilities. The new multiphase EOS moves simulations from a few-phase, single-table quantum-based representation to a more accurate many-phase, multi-table advanced quantum representation, while confirming and retaining the best and most trusted features of the current baseline EOS.
CUBIT Advances Automation of Geometry Preparation and Meshing
CUBIT is a toolkit for robust and unattended generation of two- and three-dimensional finite element meshes (grid).
To address an ASC level 2 milestone, Sandia’s CUBIT team has developed a new environment for guiding analysts through the process of generating a hexahedral or tetrahedral mesh for simulation. This new environment is called the Immersive Topology Engine for Meshing (ITEM) and is built on the existing CUBIT Geometry and Meshing Toolkit. New geometric reasoning algorithms have been developed that can detect potential problems in a CAD model and provide a list of suggested solutions. This is offered in a wizard-like environment where the user may systematically step through the geometry preparation and meshing process and run a series of diagnostic tests. The user is then presented with a list of solutions to specific geometric problems that can be easily previewed and performed at will. With phase 1 of ITEM now complete, this tool can now generate diagnostics and solutions for a range of geometric problems, including resolving small features, detecting and fixing imprint/merge problems, and detecting potential decomposition options for hexahedral sweeping. Scheduled to be completed in August 2007, this new tool promises to dramatically improve the productivity of analysts who currently must spend considerable time developing meshes for simulation.
ASC Purple Capability System Completes First Cycle Under New Usage Governance Model
Because a capability class system is similar in value and uniqueness to a large experimental facility, a governance model was set up in 2005 to allocate cycles on ASC Purple as an ongoing series of six-month campaigns. ASC Purple completed its first six months operating as a national user facility on April 16, 2007. In this first effort, ASC Purple supported three tier-one Capability Computing Campaigns (CCCs). Each campaign received significant cycles on the Purple system to meet important ASC programmatic deliverables. Technical leads for these CCCs will soon be reporting to ASC headquarters on accomplishments and lessons learned. In addition, the process to award the second round of CCC allocations on Purple is now under way. The Campaign 2 call has resulted in 20 proposals from across the three defense laboratories; the proposals are now being prioritized. ASC hopes to begin Campaign 2 on May 29, 2007.
Roadrunner Project Completes Major Milestone
The Roadrunner Project completed a major milestone in December 2006 with the Final Acceptance of the Roadrunner Base System. The Roadrunner Base System is 14 connected units (~71 teraFLOPS) for the classified computing environment and 2 connected units (~10 teraFLOPS) for the unclassified computing environment. The completion of the acceptance testing was a joint effort between IBM and Los Alamos staff.
The Roadrunner Base System is intended to provide a large “capacity-mode” computing resource for Los Alamos weapons simulations. With the completion of this milestone, on schedule, the system is now ready for additional systems integration and application readiness work in preparation for limited production in spring of 2007.
Large Dataset Generated for the Verification of an ASC Code
During March 2007 Marine Marcilhac, Division X-1 postdoc at Los Alamos (LANL), and François Hemez, X-1 technical staff member, completed 12,256 simulation runs with an ASC code to quantify numerical uncertainty. Part of the Code Verification project, headed at LANL by X-1’s Jerry Brock, the project’s goals are to develop methodologies for code and solution verification, assess time-to-solution, and quantify solution uncertainty in support of programmatic deliverables.
Read the full story…W76-1 Qualification Simulations Completed with Sandia’s Engineering Codes on Red Storm
Intense computational efforts with Sandia’s engineering codes were completed on Red Storm to support the qualification of the arming, fuzing and firing (AF&F) system for the W76-1 Life Extension Program. This effort included several diverse sets of high-fidelity simulations in which uncertainties were quantified in each application. The applications included the behavior of the AF&F under abnormal drop conditions, system behavior under normal thermal environments, radiation transport to the AF&F under hostile radiation environments, and the structural dynamics response of the AF&F to hostile environments. These applications required the use of the explicit dynamics, thermal, and structural dynamics capabilities in SIERRA Mechanics and the radiation transport capabilities in RAMSES. SIERRA Mechanics and RAMSES are two of Sandia’s engineering codes that provide mechanical/thermal/fluid and radiation/electrical capabilities, respectively. The calculations required 35% of Red Storm for more thanthree months ending in January 2007.
Tri-lab Linux Capacity Cluster Procurement Will Support Stockpile Stewardship Program
The Tri-laboratory Linux Capacity Cluster 2007 (TLCC07) procurement draft request for proposal, released in April 2007, sets the stage to supply hundreds of teraFLOPS of production capacity simulation cycles for the nation’s three defense laboratories: Lawrence Livermore, Los Alamos, and Sandia. The large-scale capacity computing platforms will be used to support the Stockpile Stewardship Program mission of the NNSA through the ASC Program. The TLCC07 clusters will be delivered to the three laboratories in stages, with the first scalable units being delivered in the fourth quarter of 2007.
Red Storm Upgrade Evaluation Receives ISC Best Paper Award
Researchers from Sandia National Laboratories were awarded one of two Best Paper awards at the 2007 International Supercomputing Conference (ISC). The paper, entitled “An Evaluation of the Impacts of Network Bandwidth and Dual-Core Processors on Scalability,” analyzes the impact of the processor and network upgrade of the Red Storm system at Sandia on ASC applications. Authors are Ron Brightwell, Keith Underwood, and Courtenay Vaughan.
Lawrence Livermore Researchers Win Best Paper Award at Computing Symposium
Lawrence Livermore researchers Bronis de Supinski and Martin Schulz, along with North Carolina State University faculty member Frank Mueller and his student Mike Noeth, won the best paper award in the software track for the 2007 International Parallel and Distributed Processing Symposium, held March 26 to 29 in Long Beach, CA.
ASC Salutes
John Turner joined Los Alamos (LANL) in 1990, in the area of Nuclear Reactor Safety and Analysis, after completing his Ph.D. in Nuclear Engineering from North Carolina State University. In 1992, he moved to X-Division to work on deterministic radiation transport in the group that would later become CCS-4. During this time, he also contributed, as a founding team member, to the casting simulation effort being initiated in the Fluid Dynamics Group, T-3. This project became Telluride, and is increasingly contributing to the improvement of manufacturing processes at LANL and elsewhere. John departed LANL, in 1997, for a stint at Blue Sky Studios, a computer animation company outside New York City.
DOE Privacy Disclaimer | Sandia Privacy Disclaimer | SAND 2007-3351W
Developed and maintained by Sandia National Laboratories for NA.114