Overview of the NNSA/ASC Labs
Total Page:16
File Type:pdf, Size:1020Kb
High Performance Computing Workshop Session VII HPC Resources and Training Opportunities at the NNSA/ASC National Labs Blaise Barney, Lawrence Livermore National Laboratory Overview • Who Are the NNSA/ASC National Labs? • Leadership in High Performance Computing • HPC Science at the Labs • Overview of HPC Platforms at LLNL, LANL and Sandia • HPC Training Opportunities at the NNSA/ASC Labs • ASC Academic Alliance Program • Future Platforms DOE CSGF HPC Workshop 1 Who Are the NNSA/ASC National Labs? • NNSA: the Department of Energy’s National Nuclear Security Administration. • Broad range of responsibilities including nuclear security, non- proliferation, naval reactors and defense programs. $24B budget • National security and stockpile stewardship in a post cold-war era are essential mission components. • ASC: Advanced Simulation and Computing Program within NNSA • Established in 1995 to support NNSA’s shift in emphasis from test based confidence to simulation based confidence. $611M budget • Under ASC, computer simulation capabilities are developed to analyze, predict and certify the performance, safety, and reliability of the nation’s nuclear stockpile. • Multi-disciplinary simulation science and implementation of the world’s most powerful computing resources. DOE CSGF HPC Workshop 2 Who Are the NNSA/ASC National Labs? • The Department of Energy funds a wide range of national laboratories and related facilities: • Ames • Los Alamos • Princeton Plasma Physics • Argonne • National Energy • Sandia • Brookhaven Technology • Savannah River • Fermi National Accelerator • National Renewable • Stanford Linear Accelerator • Idaho Energy • Thomas Jefferson National • Lawrence Berkeley • Oak Ridge Accelerator Facility • Lawrence Livermore • Three of these national laboratories are designated as NNSA defense labs: • Lawrence Livermore • Los Alamos • Sandia • The ASC program integrates the work of the NNSA defense labs to facilitate NNSA national security goals and mission. • The ASC program also funds and integrates the work of academic Alliance research centers at 8 U.S. universities (discussed later): Caltech U. Chicago U. Illinois U. Michigan Purdue Stanford U. Texas U. Utah DOE CSGF HPC Workshop 3 Who Are the NNSA/ASC National Labs? (aka Tri-labs) • Lawrence Livermore National Laboratory (LLNL) • Located in Livermore, CA; Established in 1952 • 7,800+ employees and contractors • Managed by Lawrence Livermore National Security, LLC (LLNS). Budget @ 1.6 billion • Los Alamos National Laboratory (LANL) • Located in Los Alamos, NM; Origins in the Manhattan Project (@1943) • 9,500+ employees and contractors • Managed by Los Alamos National Security, LLC (LANS). Budget @ 2.2 billion • Sandia National Laboratories (SNL) • Albuquerque, NM (main), Livermore, CA, test facilities in Nevada and Hawaii; Established 1949 • 8,400+ employees and contractors • Managed by Lockheed Martin Corporation. Budget @ 2.3 billion DOE CSGF HPC Workshop 4 Leadership in High Performance Computing The NNSA/ASC Labs have a long HPC history – even before the term "HPC" and the NNSA/ASC were invented. For example - at LLNL: DOE CSGF HPC Workshop 5 Leadership in High Performance Computing At LANL…. DOE CSGF HPC Workshop 6 Leadership in High Performance Computing At Sandia... • 1987 - 1024 node nCUBE 10 • 1990 - two 1024-node nCUBE-2’s • 1993 - Sandia fielded first Intel Paragon. 1850 nodes/3900 cpus Reached #1 on Top500 nCube • 1997-2006 - ASCI Red - Intel. World’s first Tflop system; Upgraded to 3Tflops, 4510 nodes/9298 cpus. Fastest computer on TOP500 list for 3 years. • 1997 - Cplant DEC/Compaq with Myrinet network. Achieved 997 Gflops in Nov 2003. • 2006 - current - Thunderbird - Dell. Linux cluster 65.4 Tflops. 4,480 nodes/8960 cpus. Jun 2006 #6 on TOP500 list. • 2006 - current - Red Storm - Cray. 284 Tflops theoretical peak. 12960 nodes/38,400 cpus. Nov 2009 Listed #9 in TOP500 list Paragon ASCI Red Red Storm Cplant DOE CSGF HPC Workshop 7 Leadership in High Performance Computing The NNSA/ASC Labs have dominated the Top 500 list (top500.org) for most of the 16 years since the list was created in 1993. Ranked #1 position 22 times (as of 6/09) DOE CSGF HPC Workshop 8 Leadership in High Performance Computing • Award winning science and technology • Gordon Bell Prize recognizing outstanding achievements in HPC - 16 times since it began in 1987 • Many R&D 100 Awards • Very long lists of other awards by government, industry and scientific organizations: www.sandia.gov/news/corp/awards/ www.llnl.gov/llnl/sciencetech/awards.jsp www.lanl.gov/science/awards/ • The science in the Labs’ state- of-the-art simulations is recognized in scientific publications around the world DOE CSGF HPC Workshop 9 Leadership in High Performance Computing • The NNSA/ASC Labs have been instrumental in revitalizing the U.S. HPC industry by working closely with vendors such as Cray and IBM From "Getting Up to Speed: The Future of Supercomputing" by Susan L. Graham, Marc Snir, and Cynthia A. Patterson, Editors, Committee on the Future of Supercomputing, National Research Council. 2004 Supercomputers play a significant and growing role in a variety of areas important to the nation. They are used to address challenging science and technology problems. In recent years, however, progress in supercomputing in the United States has slowed. The development of the Earth Simulator supercomputer by Japan showed that the United States could lose its competitive advantage and, more importantly, the national competence needed to achieve national goals. In the wake of this development, the Department of Energy asked the NRC to assess the state of U.S. supercomputing capabilities and relevant R&D. Subsequently, the Senate directed DOE in S. Rpt. 107-220 to ask the NRC to evaluate the Advanced Simulation and Computing program of the National Nuclear Security Administration at DOE in light of the development of the Earth Simulator. This report provides an assessment of the current status of supercomputing in the United States including a review of current demand and technology, infrastructure and institutions, and international activities. DOE CSGF HPC Workshop 10 Leadership in High Performance Computing • The NNSA/ASC Labs have been instrumental in revitalizing the U.S. HPC industry by working closely with vendors such as Cray and IBM. From "Getting Up to Speed: The Future of Supercomputing" by Susan L. Graham, Marc Snir, and Cynthia A. Patterson, Editors, Committee on the Future of Supercomputing, National Research Council. 2004 Supercomputers play a significant and growing role in a variety of areas important to the nation. They are used to address Thingschallenging have science and technology problems. In recent years, however, progress in supercomputing in the United States has changedslowed. …The development of the Earth Simulator supercomputer by Japan showed that the United States could largelylose due its competitive to advantage and, more importantly, the national competence needed to achieve national goals. In the DOE/NNSA/ASCwake of this development, the Department of Energy asked the NRC to assess the state of U.S. supercomputing capabilities investmentsand relevant inR&D. Subsequently, the Senate directed DOE in S. Rpt. 107-220 to ask the NRC to evaluate the Advanced HPC Simulation and Computing program of the National Nuclear Security Administration at DOE in light of the development of the Earth Simulator. This report provides an assessment of the current status of supercomputing in the United States including a review of current demand and technology, infrastructure and institutions, and international activities. DOE CSGF HPC Workshop 11 HPC Science at the Labs • There is a very broad range of science being explored at the Labs, much depending upon HPC: • Physics - applied, nuclear, particle & accelerator, condensed matter, high pressure, fusion, photonics • Atmosphere, Earth, Environment and Energy • Biosciences and Biotechnology • Engineering - defense technologies, laser systems (NIF), mechanical… • Chemistry • Materials • Microelectronics • Pulsed power • Computer & Information Science, Mathematics • Diverse community of scientists, researchers and collaborations - Tri-lab, other labs, universities, international… DOE CSGF HPC Workshop 12 HPC Science at the Labs: A Few Examples Concurrently using 196,608 processors in a single run, the high-fidelity simulations of a three-dimensional laser beam interacting with target is critical to achieving fusion ignition on the National Ignition Facility. BlueGene/L is refining the design of the National Ignition Facility, scheduled to achieve fusion ignition in 2010. Obtaining controlled laboratory fusion is the holy grail of national energy independence. DOE CSGF HPC Workshop 13 HPC Science at the Labs: A Few Examples DOE CSGF HPC Workshop 14 HPC Science at the Labs: A Few Examples DOE CSGF HPC Workshop 15 HPC Science at the Labs: A Few Examples University of Texas: Center for Predictive Engineering and Computational Sciences (PECOS). The goal of the PECOS Center is development of advanced computational methods for predictive simulation of multiscale, multiphysics phenomena applied to the problem of reentry of vehicles into the atmosphere. DOE CSGF HPC Workshop 16 Overview of HPC Platforms at the Tri-labs DOE CSGF HPC Workshop 17 HPC @ LLNL: BG/L and BG/P • 1999 origins in IBM exploring a novel, new architecture: massively scalable parallel with low power consumption and small footprint. • Targeted for protein folding research, but later became