InTheLoop | 03.18.2013
The Weekly Newsletter of Berkeley Lab Computing Sciences
March 18, 2013
NERSC and C3 Build the Massive Simulations Essential to Planck
To make the most precise measurement yet of the cosmic microwave background (CMB) — the remnant radiation from the Big Bang — the European Space Agency’s (ESA’s) Planck satellite mission has been collecting trillions of observations of the sky since the summer of 2009. On March 21, 2013, ESA and NASA, a major partner in Planck, will release preliminary cosmology results based on Planck’s first 15 months of data. The results have required the intense creative efforts of a large international collaboration, with significant participation by the U.S. Planck Team based at NASA’s Jet Propulsion Laboratory.
Strength in data analysis is a major U.S. contribution, including the resources of NERSC and the expertise of scientists in Berkeley Lab’s Computational Cosmology Center (C3). Read more.
ESnet’s FasterData Site Is Popular International Destination
For the past 15 years or so, ESnet’s Brian Tierney has maintained FasterData.es.net, an online repository of tips and tricks for improving network performance. Last year, 83,649 unique visitors landed at FasterData, making 120,776 visits and viewing 213,151 pages. The United States is home to more than 43,000 of those visitors, with users in the United Kingdom, India, Germany, Russia, Canada, France, China, Japan, and Brazil rounding out the top 10 countries visiting the site. Read more.
NERSC Requirements Review for Fusion Energy Sciences This Week
The NERSC Program Requirements Review “Large Scale Production Computing and Storage Requirements for Fusion Energy Sciences” is being held March 19–20 in Rockville, MD. Organized by the Department of Energy’s Office of Fusion Energy Sciences (FES), Office of Advanced Scientific Computing Research (ASCR), and NERSC, the review’s goal is to characterize FES production computing and storage requirements through 2017 at NERSC.
This review brings together DOE Program Managers, leading domain scientists, and NERSC personnel to ensure that NERSC continues to provide the world-class facilities and services needed to support DOE Office of Science research. The review will produce a report that includes HPC requirements for FES along with supporting narratives, illustrated by specific science-based case studies. Findings from the review will guide NERSC procurements and service offerings, and help NERSC, ASCR, and FES advocate for the HPC resources needed to support DOE science goals.
Sudip Dosanjh, Richard Gerber, and Harvey Wasserman are the NERSC organizers for the workshop. Alice Koniges will also participate.
ESnet User Wins CENIC’s 2013 Innovations in Networking Award
Project ADAMANT (Adaptive Data-Aware Multi-domain Application Network Topologies), a collaborative effort of the University of Southern California Information Sciences Institute (USC/ISI), the Renaissance Computing Institute (RENCI) at the University of North Carolina at Chapel Hill, and Duke University, has been honored by the Corporation for Education Network Initiatives in California (CENIC) as the recipient of the 2013 Innovations in Networking Award for Experimental/Developmental Applications.
Project ADAMANT leverages ExoGENI, a National Science Foundation-funded GENI testbed, as well as national research and education network providers of on-demand bandwidth services, specifically, National LambdaRail (NLR), Internet2 (I2), and the Energy Sciences Network (ESnet), and existing Open Science Grid computational resources to create elastic, isolated environments to execute complex distributed tasks. This approach improves the performance of these applications and, by explicitly including data movement planning into the application workflow, enables new unique capabilities for distributed data-driven “Big Science” applications. Read more.
Brent Draney and John Shalf Appear in France 24 News Video
“Twenty-four hours a day, seven days a week, servers are ready to respond to the explosion of digital information that we generate every time we use the internet. These data centres can use more electricity than a small town, and up to a whopping 90% of it is wasted. High-tech giants are now looking to put their machines on a miracle energy diet. France 24 went to California to find out more.”
So begins a news video from France 24 that includes excerpts of interviews with Brent Draney and John Shalf of NERSC — Draney discussing running an energy-efficient computing center, and Shalf discussing his research into energy-efficient computer chips. The video is available in English and French versions:
CS Researchers Contribute to APS March Meeting
The American Physical Society (APS) is holding its annual March Meeting this week, March 18–22, in Baltimore, Maryland. This is the largest physics meeting in the world, focusing on research from industry, universities, and major labs.
Contributions from Berkeley Lab Computing Sciences researchers include:
- Andrew Canning: First-Principles Electronic Structure Calculations for Scintillation Phosphor Nuclear Detector Materials (invited talk)
- Slim Chourou, Abhinav Sarje, Xiaoye Sherry Li and others: HipGISAXS: A Massively Parallel Code for GISAXS Simulation
- Jack Deslippe (co-author): An Explicit Formula for Optical Oscillator Strength of Excitons in Semiconducting Single-Walled Carbon Nanotubes: Family Behavior
- Byoungseon Jeon, Andrew Canning, and Niels Gronbech Jensen (co-authors): First Principles NEXAFS Simulations of N-Donor Uranyl Complexes
- Alexander Kemper and others: Interplay between Electron-Electron and Electron-Lattice Interactions in the RTe3 Compounds
- Lin Lin, Chao Yang, and others: The Pole Expansion and Selected Inversion Technique for Solving Kohn-Sham Density Functional Theory at Large Scale
Dan Gunter and Shreyas Cholia Giving Talk at PyData Conference
The PyData conference and workshop is a semi-annual event for scientists, engineers, and data analysts in the Python community, which is being held this Tuesday and Wednesday, March 19–20, in Santa Clara, CA. The conference focuses on techniques and tools for management, analytics, and visualization of data of different types and sizes with particular emphasis on big data.
On Wednesday, Dan Gunter, leader of CRD’s Data Intensive Systems Group, and Shreyas Cholia of NERSC’s Outreach, Software, and Programming Group will give a talk on “How Web APIs and Data-Centric Tools Power the Materials Project.” Their talk will discuss how the DOE’s Materials Project uses Python data analysis to help scientists work more collaboratively by integrating Web APIs to access remote data. They will also describe plans to extend this work to address key challenges for distributed scientific data.
CS Staff Host Albany High Students on Annual Job Shadow Day
As they have for the past five years, Computing Sciences staff members hosted Albany High School juniors as part of the school’s annual Job Shadow Day on Thursday, March 14. In all, four scientists in the Computational Research Division played host/mentor this year.
Hans Johansen and Terry Ligocki of the Applied Numerical Algorithms Group, Andy Nonaka of the Center for Computational Sciences and Engineering, and Daniela Ushizima of the Visualization Group volunteered as hosts. Students were placed according to their interests by CS Communications Manager Jon Bashor.
“It was mentally enticing to learn about the various forms of mathematical applications that exist in the world, and the computer animations and simulations amazed me no less,” wrote student Michelle Yang, who said she enjoys solving complex math problems. “Mr. Nonaka was an amiable host who made me feel especially welcome. His job descriptions were so articulate and elaborate that I feel as if I learned much more than I had anticipated. The only con, however, was the fact that the day went by too quickly!”
Adding Social Media Icons with Berkeley Lab Links to Your Gmail Signature
Berkeley Lab Public Affairs has created an online “cheat sheet” explaining how to add clickable social media icons to your Gmail signature. The web page gives instructions for adding the picture icons and links to the Berkeley Lab social media pages, but you may choose to link to the pages for Computing Sciences, NERSC, or ESnet. Here are the links to our CS social media sites:
- Berkeley Lab Computing Sciences:
- Facebook: https://www.facebook.com/LBNLcs?sk=wall
- Twitter: https://twitter.com/LBNLcs
- Google+: https://plus.google.com/u/0/101212778950573946812/posts
This Week’s Computing Sciences Seminars
Achieving Performance Portability on HPC Systems
Monday, March 18, 10:00–11:00 am, 50F-1647
Pei-Hung Lin, University of Minnesota
The past decade has produced numerous CPU architectural innovations. These have included multiple cores per CPU, multiple simultaneous threads per core, and, especially with GPUs, highly complex memory hierarchies. As a result, achieving high performance has become a major challenge to programmers. The SIMD engines and on-chip memory in modern CPU and GPU cores are identified as the keys to obtaining high performance for scientific application codes. These common elements in present computing devices make performance portability possible.
PPM gas dynamics applications have delivered high percentage of peak performance on latest HPC systems. Achieving such high performance for CFD applications using structured grids requires programmers to express the code in an optimized form. An automated approach that generates the optimized code can greatly relieve programmer’s burden and deliver performance portability.
In this talk, I will present the optimization strategies employed in the domain-specific tool developed by my team for the PPM gas dynamic applications. These strategies deliver performance portability on current and future HPC systems.
Building Text-Analysis Tools for Literary Study
Monday, March 18, 12:00–1:00 pm, Geballe Room, Stephens Hall, UC Berkeley
Marti Hearst and Bryan Wagner, UC Berkeley
Marti Hearst is a professor in the School of Information at UC Berkeley, with an affiliate appointment in the Computer Science Division. Her primary research interests are user interfaces for search engines, information visualization, natural language processing, and empirical analysis of social media. Hearst has recently completed a book on Search User Interfaces. Bryan Wagner is associate professor in the English Department and affiliated faculty in the American Studies Program at UC Berkeley. His primary research focuses on African American expression in the context of slavery and its aftermath, with secondary interests in legal history, cultural theory, and popular music.
Hearst and Wagner will discuss their National Endowment for the Humanities funded WordSeer project. This project began as an experiment by Ph.D. candidate Aditi Muralidharan asking how natural language processing and data visualization could be applied to the process of literary study. Two years and two NEH grants later, the team has expanded their goal from matching up computational linguistics algorithms with literary questions to the creation of a publicly available scholarly tool for the visual and algorithmic analysis of text.
A brown bag lunch conversation with Hearst and Wagner discussing lessons learned and the close collaboration, field-testing, and detailed understanding of humanistic work processes that are necessary for tool builders to be successful.
A Floating-Point Trick to Solve Boundary-Value Problems Faster: Scientific Computing and Matrix Computations Seminar
Wednesday, March 20, 12:10–1 pm, 380 Soda Hall, UC Berkeley
William Kahan, UC Berkeley
This talk resuscitates an old trick to accelerate the numerical solution of certain discretized boundary-value problems. Without the trick, half the digits carried by the arithmetic can be lost to roundoff when the discretization’s grid-gaps get very small. The trick can obtain adequate accuracy from arithmetic with float variables 4 bytes wide instead of double variables 8 bytes wide. Wider data moves slower through the computer’s memory system and pipelines. The trick is tricky for programs written in MATLAB 7, Java, Fortran and post-1985 ANSI C. The trick is easy for the original Kernighan-Ritchie C of the late 1970s, and for a few implementations of C99 that fully support IEEE Standard 754 for Binary Floating-Point.
NERSC Brown Bag: Revolution Analytics
Thursday, March 21, 11:00 am–1:00 pm, OSF 943-236
Dani Ushizima, LBNL/CRD
Adiabatic Quantum Annealing
Thursday, March 21, 1:00–2:00 pm, 50B-4205
Robert F. Lucas, Director of Computational Sciences, Information Sciences Institute (ISI), University of Southern California
The USC–Lockheed Martin Quantum Computing Center has taken delivery of a D-Wave adiabatic quantum annealing device. This talk will report on our experience assessing the quantum mechanical behavior of the device. We will present experimental results that strongly indicate that quantum annealing is indeed being performed by D-Wave. We will also discuss benchmarking of its performance relative to classical systems, and ongoing efforts to develop applications for it.
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.