July 29, 2013
The Keeneland Project (a five-year Track 2D cooperative agreement awarded by the National Science Foundation (NSF) in 2009 for the deployment of an innovative high performance computing system in order to bring emerging architectures to the open science community.) has been credited with helping researchers analyze cancer tumors. Read the article HERE.
Modeling and Simulation Made NiCE
May 4, 2013
The NEAMS Integrated Computational Environment - NiCE for short - is the NEAMS program's answer to the need for easy usability, accessibility, and collaboration. It complements both SHARP and MOOSE from the reactors and fuels product lines with common tools for creating input files, launching jobs locally and remotely, looking at data in 3D, and managing "assets," such as simulation input and output files. [more]
CLIMATE - Going small with big computers . . .
ORNL's supercomputers are allowing climate scientists to zoom in on smaller and smaller areas to try to determine whether local and regional droughts influence climate extremes on a larger scale. What is the probability of a drought developing in the Southwest this decade? Ultra-high-resolution climate models may improve our ability to provide informed projections. In the highest-resolution model, grid cells are a mere one-quarter of a degree (23 miles) wide. "Data at this scale is only accessible with leadership computing resources," says ORNL mathematician Rick Archibald, one of nine ORNL researchers collaborating with scientists at Lawrence Berkeley, Lawrence Livermore and Los Alamos national laboratories on the Ultra-high-resolution Global Climate Simulation project.
A New Era in Automotive Engine Development Driven by HPC
CSMD researcher Sreekanth Pannala was interviewed about high performance computing's role in the development of new engines. [more]
CSMD researchers Bo Liu and Yehuda Braiman (along with ORNL researcher Yun Liu) have been awared a patent for their invention: V-shaped Resonators for addition of broad-area laser diode arrays. This invention is drawn to systems and methods for coherent addition of broad-area laser diode array.
Full patent document [HERE]
CSMD researcher Moetasim Ashfaq's paper "Response of snow-dependent hydrologic extremes to continued global warming" has been published by Nature magazine's online presense Nature Climate Change. [more]
High Performance Computing Key Enabler for Accelerating Development of High Efficiency Engines
The complexity of new and future vehicles - driven by the need for increasing fuel efficiency and decreasing emissions with ever-changing drive-cycle demands and environmental conditions - is adding unprecedented flexibility in design and driving the need for better simulation and more powerful computers, observed Dr. Robert M. Wagner, Director of the Fuels Engines and Emissions Research Center, and Dr. Sreekanth Pannala, Senior Research Staff Member in the Computing and Computational Sciences Directorate at Oak Ridge National Laboratory in a keynote talk at the recent Global Powertrain Conference. [more]
CSMD Researcher Forrest Hoffman (along with ORNL researchers Jitendra Kumar and Richard Mills), is on the team that received the United States Department of Agriculture's 2012 Director's Science Delivery Award for their work on the USDA's project ForWarn.
ForWarn is a satellite-based forest disturbance monitoring system for the conterminous United States. It delivers new forest change products every eight days and provides tools for attributing abnormalities to insects, disease, wildfire, storms, human development or unusual weather. Archived data provide disturbance tracking across all lands since 2000. http://forwarn.forestthreats.org/
Novel capability will deliver the best of high-performance computing and cloud computing at national facility
Next-generation neutron scattering requires next-generation data analysis infrastructure. And that means not just more data, accelerated reduction, and translation and analysis, but linking the neutron scattering on a beam line live to a simulation platform where modeling and simulation can guide the experiment. [more]
David Bernholdt was named Group Leader of the Computer Science Research Group (CSR) on November 1. CSR represents the merger of the Application Performance Tools group, previously led by Rich Graham and CSR previously led by Al Geist. Al, who led the CSR group for 20 years, will be spending more time in his role as Chief Technology Officer for NCCS and in other leadership roles within CSMD. The mission of the CSR group helps researchers create and run scientific and technical applications on the largest scales of HPC systems. The group focuses on extreme-scale systems; system software; program- ming environments, including compilers and other program translation approaches, runtimes and communications, debug- gers, performance, and other tools; and scientific software. In 1993 he received his Ph.D. in Chemistry from the University of Florida, and then spent 2 years as a postdoc at PNNL as part of the original development team for the NWChem parallel com- putational chemistry package. From 1995-2000, he was an Alex G. Nason Fellow & a Sr. Research Scientist at Syracuse Universi- ty, and carried out research in both computational chemistry and computer science. Then he joined ORNL where his research program has focused on programming environments for HPC.
CSMD researcher Rahul Sampath's mentee, Damien Lebrun-Grandie, won second place in the 2012 Oak Ridge National Laboratory Nuclear Engineering Science Laboratory Synthesis (NESLS) Poster session. Twenty-eight of the NESLS students presented their work at the 2012 Summer Student Poster Session, which were judged for their excellence in technical achievement and presentation. The participants represent many areas of research throughout the laboratory, including research reactors, safeguards and nonproliferation, material sciences, nonreactor nuclear facilities, reactor and nuclear systems, radioisotope production, and CASL.
CSMD researcher Rahul Sampath is a co-author on the paper A Massively Parallel Adaptive Fast Multipole Method on Heterogeneous Architectures. This paper was invited to be a Research Highlight for the publication Communications of the ACM. The paper describes a parallel fast multipole method (FMM) for highly nonuniform distributions of particles. We employ both distributed memory parallelism (via MPI) and shared memory parallelism (via OpenMP and GPU acceleration) to rapidly evaluate two-body nonoscillatory potentials in three dimensions on heterogeneous high performance computing architectures. Click HERE to read the full article.
Thomas R. O'Donnell - DEIXIX Magazine
With the help of Oak Ridge computations, scientists are probing the properties of macroscale sponges made of nanoscale carbon-boron tubes. The material could soak up oil spills, help store energy or meet other needs.
In 2007, when Oak Ridge National Laboratory (ORNL) researchers calculated that adding boron would bend carbon nanotubes, they did little with the information.
Boron was one of several elements the computational scientists plugged into their model as they investigated ways to induce useful changes in nanotube structures. There were experiments to compare with the results of most of their calculations. There weren't any to check against the boron-doped nanotube simulations.
"We didn't think anything about boron, really," says Bobby Sumpter, Chemical and Materials Sciences Group leader and director of ORNL's Nanomaterials Theory Institute. "We thought it was interesting how it preferred negative curvature, and we kind of just left it at that."
Then Humberto Terrones, an ORNL-affiliated researcher from Belgium's Université Catholique de Louvain, came to visit last year. He and his brother, Mauricio, of Pennsylvania State University and Japan's Shinshu University, were investigating new nanotube materials.
Humberto Terrones "was talking about how they'd observed these three-dimensional-looking structures when they doped boron in," Sumpter recalls. "I said, 'But, Humberto, remember our results? Where we found these interesting effects and we think we understand exactly what happens?' I hadn't realized they'd done experiments for boron and just learned about it over a casual discussion - which actually turn out to be usually the most productive scientific discussions - just a cup of coffee with a white board."
Sumpter and Vincent Meunier of Rensselaer Polytechnic Institute recalculated the boron results and published them jointly with Rice University doctoral student Daniel Hashim's discovery of three-dimensional, macro-scale nanosponges. "It's always good to have experimental evidence that backs up theory or vice versa," Hashim says. "In this case we made the theory and the experimental evidence together and it gave this paper a lot more impact."
Click HERE for more information.
Student Wins ACM Student Research Competition
Auburn University student Yuan Tian won first place in the graduate category at the Association for Computing Machinery's (ACM) student research competition for her paper "Smart-IO: System-Aware Two-Level Data Organization for Efficient Scientific Analytics." Yuan is currently working in CSM's Scientific Data Group which is led by Dr. Scott Klasky. Dr. Klasky was pleased to point out that Yuan "has been integral for our continued success with ADIOS, and I believe she will be an incredible success in the future, as a postdoctoral scholar in the NSF RDAV project."