The US Department of Energy has allocated about 10.4 million CPU hours on supercomputers at the National Energy Research Scientific Computing Center (NERSC) at Lawrence Berkeley National Laboratory as part of a program to accelerate scientific discoveries in multiple disciplines, including climate, physics, combustion and material science.
The one-year allocations will go to 11 projects by researchers in universities, national labs and industry. Last year, the DOE allotted nearly 9 million CPU hours at NERSC to seven projects.
The awards are part of a program called Innovative and Novel Computational Impact on Theory and Experiment (INCITE), launched in 2003.
INCITE, supported by the DOE Office of Science, selects projects that require large-scale and intensive use of supercomputers and also promise to deliver a significant advance in science and engineering.
Overall, the INCITE program is awarding more than 265 million CPU hours to 55 projects for 2008, up from 95 million CPU hours for 45 projects in 2007.
Raymond L. Orbach, DOE under-secretary for Science, said: ‘The Department of Energy's Office of Science has two of the top 10 most powerful supercomputers, and using them through the INCITE program is having a transformational effect on America's scientific and economic competitiveness.’
Orbach added: ‘Once considered the domain of only small groups of researchers, supercomputers today are tools for discovery, driving scientific advancement across a wide range of disciplines. We're proud to provide these resources to help researchers advance scientific knowledge and understanding and thereby to provide insight into major scientific and industrial issues.’
In addition to the projects at NERSC, other INCITE projects were awarded time at DOE's Leadership Computing Facilities at Oak Ridge National Laboratory in Tennessee and Argonne National Laboratory in Illinois, and the Molecular Science Computing Facility at Pacific Northwest National Laboratory in Washington. As the flagship facility for the Office of Science, NERSC provided the only computing resources available during the first two years of the program.
One of the projects at NERSC is led by Gilbert Compo from the University of Colorado who will produce a global tropospheric circulation dataset dating back to 1892. The dataset will help validate the climate models being used to make climate projections for the 21st century. The only dataset available for the early 20th century consists of error-ridden, hand-drawn analyses of the mean sea level pressure field over the Northern Hemisphere.
‘The allocation has been invaluable. Without it, we could not have generated a dataset of the six-hourly global weather maps spanning 1918 to 1949 that will be used to understand the Dust Bowl and dramatic Arctic warming of 1920-1940s, among other climate and weather anomalies of the period," said Compo, who also works with the National Oceanographic and Atmospheric Administration (NOAA). Compo, who received an INCITE allocation last year, will be computing at NERSC again this year.
The other 10 projects awarded computing time at NERSC are:
John Bell from Berkeley Lab will lead a computational study to enable a fundamental understanding and characterisation of thermo-diffusively unstable flames in both atmospheric and high-pressure regimes relevant to ultra-lean turbulent premixed burners. The research will aid the development of near-zero-emission combustion devices, a goal of the FutureGen power plant project sponsored by the DOE's Office of Fossil Energy.
Hong Im from the University of Michigan will lead the work on developing three-dimensional simulations of turbulent non-premixed flames in the presence of a mean flow strain and fine water droplets. The simulations will help address important issues on energy and environmental research on mixed burners. The research will aid the development of near-zero-emission combustion devices, a goal of the FutureGen power plant project sponsored by the DOE's Office of Fossil Energy.
Warren Washington from the National Center for Atmospheric Research (NOAA) will continue the development of the Climate Science Computational End Station (CCES), models for simulating and predicting climate change. The research will examine the human impact on the climate and improve the accuracy of climate models, including the simulation of the global carbon cycle.
David Randall from Colorado State University will head a project to simulate the global circulation of the atmosphere with roughly a two-kilometre grid spacing. Understanding the role of clouds in the global atmosphere is key to developing more accurate climate models.
Warren Mori from the University of California at Los Angeles will develop simulations to answer questions about plasma-based particle accelerators that currently cannot be answered through experiments. New acceleration techniques using lasers and plasmas could lead to ultra-compact accelerators.
Chuang Ren from the University of Rochester will carry out large-scale particle-in-cell (PIC) simulations of the ignition phase in fast ignition (FI), one of the most promising new methods for improving the viability of inertial confinement fusion as a practical energy source.
Lawrence Pratt at Fisk University plans to examine several lithium compounds that are among the best reagents for forming carbon-carbon bonds in organic synthesis, which can lead to the development of powerful medicines. The project will use ab initio and density functional theory methods to investigate the structure and reactions of the organolithium compounds.
Ji Qiang from Berkeley Lab will optimize the design and improvement of beam delivery systems for the next-generation X-ray free electron lasers (FELs), which have excellent applications in physics, material science, chemical science and bioscience. Optimising the beam delivery systems to produce and preserve high intensity and good quality electron beams will not only lower the cost of design and operation of FELs, but also improve the performance of the X-ray light output.
Leeor Kronik from the Weizmann Institute of Science in Isarel will use the many body perturbation theory in understanding the structures of novel electronic materials. The results from the research will help clarify pressing issues in figuring out the electronic structure of organic/inorganic interfaces with applications in areas such as semiconductors.
A project by Fluent, in partnership with General Motors, will use its software to perform computational fluid dynamics and thermal calculations for designing cars. The research will tackle five areas, including the full-vehicle open sunroof wind buffeting calculations and the simulations of semi-trucks passing stationary vehicles with raised bonnets.