Sandia National Laboratories Parallel Supercomputing Achievements
Last modified: August 6, 1997
Questions and Comments ||
Privacy and Security
Sandia scientists win the first two national awards given for parallel computing: the Karp Challenge to demonstrate a more than 200-fold speed increase through computer parallelism, and the inaugural Gordon Bell Award for contributions to parallel computing. Sandia used the nCUBE 10 computer to achieve near perfect speedup on software for solving real problems in structures and acoustics. The software wins an R&D 100 Award.
Sandia develops the CTH codes, used for computer modeling of high-speed impacts and the effects of nuclear explosions. The latest version of the codes, PCTH, uses the immense power of massively parallel supercomputers to effectively simulate complicated systems requiring massive datasets, and has been used to help resolve nuclear weapons safety problems. The code is selected as the Department of Defense code of choice for solving high velocity impact problems.
Sandia scientists develop massively parallel algorithms and models to describe solid state, chemical, and biological systems on length scales ranging from microscopic (systems containing hundreds of atoms) to macroscopic (systems containing millions of atoms). The goal is to provide the Department of Energy and industry with a computer design capability for materials and molecules, and these codes have been applied successfully in several instances.
Sandia and University of New Mexico researchers develop SUNMOS (Sandia-University of New Mexico Operating System), doubling the memory and throughput on Intel Paragon computers worldwide, and forming the basis of a cooperative research and development agreement with Oracle and nCUBE. The operating system is in use in about two dozen sites worldwide.
A new Sandia-developed software package, CHACO, provides the means to make massively parallel computers easier to use by facilitating the mapping process that distributes computations across multiple processors. This improves efficiency of a massively parallel supercomputing by helping to ensure each processor has about the same amount of work to do and that the quantity of interprocessor communication is limited. CHACO is licensed to over 150 sites around the world.
Sandia researchers use supercomputers to create detailed, three-dimensional medical images with data collected through magnetic resonance imaging (MRI). Special computer codes animate the MRI images, allowing physicians to better assess lesions or other abnormalities.
Sandia participates in a very successful interdisciplinary collaboration with the Johns Hopkins School of Medicine to study the detailed biochemistry of the active metabolite phosphoramide mustard (PM). This effort has produced a detailed understanding of aqueous phase properties of PM and some results about the configuration of the DNA bound structure.
Sandia uses massively parallel computing to perform real time man-in-the-loop simulations, which are part of submarine tactical exercises. The technique is dubbed stimulation rather than simulation because it exercises actual war fighting hardware and personnel. The technique will be applied to surface ship exercises in 1995.
Sandia scientists use the PCTH family of computer codes and the Paragon massively parallel computer to accurately simulate the collision of Shoemaker-Levy 9 comet fragments with Jupiter. The cometís discoverers credit Sandia for alerting the astronomical community that the impact would cause fireballs visible from Earth. The work wins the Best Paper Award from the Hypervelocity Impact Society.
A Sandia-led team wins the labs' second Gordon Bell Award, achieving record performance and showing that large, scalable machines are practical for commercial applications and can give companies significant market advantage. The team used a parallel dense equation solver in four commercial applications: structural analysis, acoustic radiation scattering, electromagnetic radiation modeling, and high-frequency ray tracing.
Sandia scientists develop Salsa, a program designed to solve a broad range of chemically reacting flow problems, including combustion problems for transportation, atmospheric chemistry modeling for pollution studies, and other reacting flows in manufacturing processes such as chemical vapor deposition, a key step in semiconductor manufacturing. Salsa computes the solution of the governing transport equations for momentum, total mass, and thermal energy, together with individual gas- and surface-phase chemical species for laminar flows, and is being used in the semiconductor industry.
Sandia and Intel regain the world record for computational speed by achieving 281 billion floating point operations per second.
Sandia's massively parallel quantum chemistry program is applied to several real-world problems in medicinal chemistry including studies of anticancer drugs and environmental carcinogens. This software is also being applied to chemical simulations in support of other DOE programs including the detoxification of organophosphate nerve gases, the energy conversion of chemical explosives, and the development of biosensors for chemical and biological warfare agents.
More than two dozen engineering and science codes are running on Sandia massively parallel machines. Sandia uses PCTH to help resolve an important safety issue related to fragmentation.
Sandia scientists use massively parallel machines to create three-dimensional models simulating underground formations for the gas and oil industry. The simulations allow for the rapid identification of oil deposits in a matter of hours, a task that previously took a month or more. The work is part of a cooperative research and development agreement.
Sandia develops a new molecular dynamics code that is suitable for modeling large molecular systems, an important task in drug-design analyses, as part of a cooperative research and development agreement with Bristol-Myers Squibb, Cray Research, and Dupont.
Sandia continues to refine computational models for predicting material property changes in weapon components due to aging. Once teraflop computing capability is available, the current lattice code for modeling microcracking in materials will enable scientists to predict material deterioration and thus prevent component failures in the weapons stockpile.
Sandia scientists develop QUEST (QUantum Electronic STructure), an electronic structure program designed for massively parallel supercomputers, although it is portable and will run in a workstation environment. The program is useful for understanding systems at a quantum level and is important for a variety of applications such as catalysis, drug design, microelectronics, new chemicals, and new alloys.
Intel computer engineers demonstrate the one-teraflops breakthrough, considered to be the Holy Grail of high-performance computing, to officials from the Department of Energy and Sandia National Laboratories on Dec. 11 in Beaverton, Ore. The Intel massively parallel computer was developed under direction of the DOE for the Accelerated Strategic Computing Initiative (ASCI), a 10-year program designed to move nuclear weapons design and maintenance from a test-based to simulation-based approach.
Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy.
Larry Perrine, email@example.com (505) 845-8511
Back to top of page || Sandia Home Page
Last modified: August 6, 1997
Questions and Comments || Privacy and Security