InTheLoop | 04.25.2011
April 25, 2011
NERSC and ESnet Launch New Web Sites
ESnet has launched a new web site at www.es.net, and NERSC is scheduled to launch a new site later today at www.nersc.gov. The new Web 2.0 versions were a year in the making. Both sites feature a modern look and feel, easier navigation, and improved search.
The ESnet site features a Network Weathermap that shows current network traffic conditions. The NERSC site has a new “For Users” section designed to help NERSC users easily find what they are looking for, and a new “My NERSC” page that gives a convenient summary of job and usage information.
Supercomputer Cracks “Impossible” Calculation
The BlueGene/P supercomputer system, used for IBM’s benchmarking tests and quality control, has been used to conquer a calculation thought to be unachievable. The result is the paper “The Computation of Previously Inaccessible Digits of π2 and Catalan’s Constant” by CRD Chief Technologist David H. Bailey, Jonathan M. Borwein of Newcastle University in Australia, and Andrew Mattingly and Glenn Wightwick of IBM Australia.
Bailey said the project came about when IBM Australia was looking to do something related to “Pi Day” (March 14) on a new IBM BlueGene/P computer system. Bailey’s colleague Jon Borwein suggested that IBM use one of the formulas (originally discovered in part by Bailey) that permit one to calculate digits of pi squared, beginning at some huge starting point (this has already been done for pi itself, also using formulas discovered in part by Bailey). So beginning with a computer program originally written by Bailey, IBM researchers developed highly efficient parallel code to compute digits of both pi squared and also Catalan’s constant, another important number that arises in mathematics and mathematical physics.
In the end, the IBM researchers did three rather stupendous calculations — pi squared in binary (beginning at the 60 trillionth binary digit), pi squared in base-3 (beginning at the 60 trillionth base-3 digit), and Catalan’s constant in binary (beginning at the 120 trillionth binary digit).
“What is interesting in these computations is that until just a few years ago, it was widely believed that such mathematical objects were forever beyond the reach of human reasoning or machine computation,” Bailey said. “Once again we see the utter futility in placing limits on human ingenuity and technology.”
Horst Simon Is Guest on Radio Show About Humans vs. Computers
When the IBM computer Watson snatched the Jeopardy! title from its human competition, it raised the question of how smart machines are. Could artificial intelligence ever beat humans at being human? During a recent radio program, a panel of experts — including Berkeley Lab Deputy Director Horst Simon — discussed making peace with your PC, the self-evolution of technology, and whether machines will ever beat the human mind. More>
Andy Nonaka to Speak at Bay Area Scientific Computing Day
Andy Nonaka of CRD’s Center for Computational Sciences and Engineering will present a talk on “Petascale Simulations of Type Ia Supernovae” at the Bay Area Scientific Computing Day (BASCD) on Sunday, May 8, at Stanford University. Presented by Stanford’s Institute for Computational and Mathematical Engineering (ICME), BASCD is an annual informal gathering to encourage the interaction and collaboration of researchers in the fields of scientific computing and computational science/engineering from the San Francisco Bay Area. This event provides a great venue for junior researchers to present their work to the local community, and for the Bay Area scientific and computational science/engineering communities at large to interchange views on today’s multidisciplinary computational challenges and state-of-the-art developments.
ICME Open Day on Saturday, May 7, will be a day of talks, demos and posters showcasing some of the exciting research projects that faculty and students of ICME are involved in. Agendas and registration for both events are available on their web pages.
MSRI Presents Staged Play Reading of “MSI: Anatomy”
The Mathematical Sciences Research Institute (MSRI) and the UC Berkeley Mathematics Department are presenting a special performance of Andrew Granville’s “MSI (Mathematical Science Investigation): Anatomy of Integers and Permutations” in a staged play reading on Friday, April 29, 7:00–8:30 pm in MSRI’s Simons Auditorium. A Q&A session with Andrew Granville will follow the presentation.
“MSI: Anatomy” is an experimental work that blurs the boundaries between pure mathematics, film, and live performance. Andrew Granville, mathematician and vulgarizer; Jennifer Granville, actor and screenwriter; and Michael Spencer, performance designer, have collaborated to present this special rehearsed reading at MSRI. Thrill to mysterious murders, marvel at detectives’ deductions, and groan at the mathematical puns—for one night only!
You will discover that the “MSI: Anatomy” plot thickens: There have been two homicides. The bodies are in the morgue and are ready to be examined. It does not seem like there is any link between the two crime victims, an integer and a permutation. Mathematical forensic experts are ready to dissect the corpses and to study the anatomy of the victims. There will be several (mathematical) surprises. Come and participate in the autopsy!
Admission is free, but seating is very limited; to reserve a seat you must register here. If you are bringing a guest, please register them by entering their name and using your email address. It is highly unlikely that any seats will be available on the night of the play, so please sign up in advance.
This Week’s Computing Sciences Seminars
Dremel: Interactive Analysis of Web-Scale Datasets
Monday, April 25, 11:00 am–12:00 pm, 405 Soda Hall, UC Berkeley
Matt Tolton, Google Inc.
Dremel is a scalable, interactive ad-hoc query system for analysis of read-only nested data. By combining multi-level execution trees and columnar data layout, it is capable of running aggregation queries over trillion-row tables in seconds. The system scales to thousands of CPUs and petabytes of data, and has thousands of users at Google. I’ll describe the architecture and implementation of Dremel, and talk about how it complements MapReduce-based computing. I’ll present a novel columnar storage representation for nested records and discuss experiments on few-thousand node instances of the system.
Navigating NERSC File Systems
Tuesday, April 26, 10:00–11:00 am, OSF 943-238
Please register to participate online
David Turner, NERSC User Services Group
NERSC hosts a number of file storage systems, each with its unique characteristics. In this tutorial we will describe best uses for each file system, their data retention policies, how to move data among them, and how to share data with colleagues.
MATLAB/Simulink Seminar: Data Analysis and Visualization in MATLAB
Tuesday, April 26, 10:00 am–12:00 pm, Bldg. 50 auditorium
Doug Eastman, Senior Application Engineer, MathWorks
In this session, we will demonstrate how to acquire, analyze, and visualize data through mathematical, statistical, and engineering functions that support common engineering operations. The session also provides an overview of the MATLAB technical computing environment. Highlights include:
- Importing data from various sources
- Performing statistical analysis and curve fitting
- Automating analysis via MATLAB code generation
- Building GUIs and generating reports
Par Lab Seminar: Architectural Musings: Rethinking Mobile Computer Systems Architecture
Tuesday, April 26, 11:00 am–12:30 pm, Soda Hall, Wozniak Lounge, UC Berkeley
Christopher Vick, Qualcomm
Modern computer systems are complex combinations of hardware and software. These systems have evolved over many years, and their basic structure is derived from a fundamental body of work in computer science and computer architecture that was developed many years ago. This body of work focused on optimizing the utilization of scarce resources, and working around key system bottlenecks in order to extract the most value possible from computer systems. In modern mobile computer systems, the set of resources which are scarce and the set of system bottlenecks has changed quite dramatically from those older systems. This talk explores some of the changes in scarce resources and system bottlenecks, and how those changes might affect the design of mobile computer systems in the future.
MATLAB/Simulink Seminar: How a Differential Equation Becomes a Robot
Tuesday, April 26, 1:00–4:00 pm, Bldg. 50 auditorium
Doug Eastman, Senior Application Engineer, MathWorks
This seminar will show how the suite of MathWorks tools complement and enhance each other, and how when combining them together, the user can unleash the full potential of our complete development environment. Starting from underlying mathematical and physical principles, we will discuss the iterative process of analysis, design and optimization involved in the development and implementation of a real-world practical application. The demonstration example will examine how a simple second order differential equation can evolve into a complex dynamic model of a multi-degree of freedom robotic manipulator that includes the controls, electronics and three-dimensional mechanics of the complete system. Highlights of the presentation include:
- Using the MuPad interface in the Symbolic Math Toolbox to create equations of motion
- Modeling complex electro-mechanical systems using Simulink and the physical modeling libraries
- Using MATLAB scripts to control and automate the simulation of dynamic models in Simulink
- Importing three-dimensional mechanisms directly from CAD packages using the SimMechanics translator
- Using the Control System Toolbox and the Optimization Toolbox directly on your Simulink model
- Creating virtual reality visualizations and interfacing with external devices using Simulink 3D Animation
- Prototyping and testing your real-time system directly in hardware with xPC Target
Numerical Solvers for Nonlinear Partial Differential Equations on Octree Adaptive Grids
Tuesday, April 26, 3:00–4:00 pm, 50B-2222
Frederic Gibou, University of California, Santa Barbara
Several phenomena in the physical and the life sciences can be modeled as a time dependent interface problem and nonlinear partial differential equations. Examples include the study of electro-osmotic flows, molecular beam epitaxy, free surface flows and multiphase flows in porous media. One of the main difficulties in solving numerically these equations stems from the fact that the geometry of the problems is often arbitrary and special care is needed to correctly apply boundary conditions. Another difficulty is associated with the fact that such problems involve dissimilar length scales, with smaller scales influencing larger ones so that nontrivial pattern formation dynamics can be expected to occur at all intermediate scales. Uniform grids are limited in their ability to resolve small scales and are in such situations extremely inefficient in terms of memory storage and CPU requirements. In this talk, I will present recent advances in the numerical treatment of interface problems and describe new numerical solvers for nonlinear partial differential equations in the context of adaptive mesh refinement based on Octree grids. If time permits, I will present a second-order accurate symmetric positive definite monolithic solver for fluid/solid interactions.
LAPACK Seminar: Numerical Solvers for Nonlinear Partial Differential Equations on Octree Adaptive Grids
Wednesday, April 27, 11:10 am–12:00 pm, 380 Soda Hall, UC Berkeley
Frederic Gibou, UC Santa Barbara
Several phenomena in the physical and the life sciences can be modeled as a time dependent interface problem and nonlinear partial differential equations. Examples include the study of electro-osmotic flows, molecular beam epitaxy, free surface flows and multiphase flows in porous media. One of the main difficulties in solving numerically these equations stems from the fact that the geometry of the problems is often arbitrary and special care is needed to correctly apply boundary conditions. Another difficulty is associated with the fact that such problems involve dissimilar length scales, with smaller scales influencing larger ones so that nontrivial pattern formation dynamics can be expected to occur at all intermediate scales. Uniform grids are limited in their ability to resolve small scales and are in such situations extremely inefficient in terms of memory storage and CPU requirements. In this talk, I will present recent advances in the numerical treatment of interface problem and describe new numerical solvers for nonlinear partial differential equations in the context of adaptive mesh refinement based on Octree grids. If time permits, I will present a second-order accurate symmetric positive definite monolithic solver for fluid/solid interactions.
Biomedical Applications of Dynamic Quantitative Imaging: Oncology, Developmental Biology, and Neuroscience
Friday, April 29, 11:00 am–12:30 pm, 521 Cory Hall (Hogan Room), UC Berkeley
Thomas M. Baer, Executive Director, Stanford Photonics Research Center, Stanford University
Our understanding of the molecular basis of life has been greatly expanded by the development of highly precise analytical instruments capable of measuring tens of thousands of molecular targets simultaneously. In parallel with these revolutionary developments, new imaging technologies have also rapidly evolved, providing unsurpassed resolution and highly accurate three dimensional images of cells and organ systems in living organisms. A growing research frontier is at the intersection of these two areas: correlating dynamic quantitative imaging data with precise multi-parameter molecular analysis of DNA, RNA, and proteins in vivo, and from samples harvested from live tissue. Working at this intersection requires mathematical modeling of complex molecular data sets and development of automated image analysis and feature extraction algorithms of large three dimensional image data sets. In this seminar I will discuss several examples of recent research using this approach in studies in human development biology, neuroscience, and oncology.
Link of the Week: Is Sitting All Day Killing You?
An article in Men’s Health says that regardless of how often or how hard you work out, there’s still a good chance that you’re sitting your life away. The more hours a day you sit, the greater your likelihood of dying an earlier death regardless of how much you exercise or how lean you are. “Nonexercise activity” —standing or walking — is an independent health factor, especially for cardiovascular disease.
In “Sitting Time and Mortality from All Causes, Cardiovascular Disease, and Cancer,” published in the journal Medicine & Science in Sports & Exercise, P. T. Katzmarzyk et al. concluded:
These data demonstrate a dose-response association between sitting time and mortality from all causes and CVD [cardiovascular disease], independent of leisure time physical activity. In addition to the promotion of moderate-to-vigorous physical activity and a healthy weight, physicians should discourage sitting for extended periods.
About Computing Sciences at Berkeley Lab
The Lawrence Berkeley National Laboratory (Berkeley Lab) Computing Sciences organization provides the computing and networking resources and expertise critical to advancing the Department of Energy's research missions: developing new energy sources, improving energy efficiency, developing new materials and increasing our understanding of ourselves, our world and our universe.
ESnet, the Energy Sciences Network, provides the high-bandwidth, reliable connections that link scientists at 40 DOE research sites to each other and to experimental facilities and supercomputing centers around the country. The National Energy Research Scientific Computing Center (NERSC) powers the discoveries of 6,000 scientists at national laboratories and universities, including those at Berkeley Lab's Computational Research Division (CRD). CRD conducts research and development in mathematical modeling and simulation, algorithm design, data storage, management and analysis, computer system architecture and high-performance software implementation. NERSC and ESnet are DOE Office of Science User Facilities.
Lawrence Berkeley National Laboratory addresses the world's most urgent scientific challenges by advancing sustainable energy, protecting human health, creating new materials, and revealing the origin and fate of the universe. Founded in 1931, Berkeley Lab's scientific expertise has been recognized with 13 Nobel prizes. The University of California manages Berkeley Lab for the DOE’s Office of Science.
DOE’s Office of Science is the single largest supporter of basic research in the physical sciences in the United States, and is working to address some of the most pressing challenges of our time. For more information, please visit science.energy.gov.