InTheLoop | 12.03.2012
December 4, 2012
Climate Change Study Strengthens Link to Human Activities
New research shows some of the clearest evidence yet of human influence on atmospheric temperature. Published online in the November 29 early edition of the Proceedings of the National Academy of Sciences, the research compares data from climate models and satellite observations over a 33-year period. Read more.
Phase 1 of Edison Arrives at NERSC
Phase 1 of NERSC’s newest supercomputer, named Edison, was delivered on November 27, 2012. The architecture is a Cray XC30 ("Cascade") and it will be installed in two phases. When it is fully installed in 2013, Edison will have a peak performance of more than 2 petaflops (1015 floating point operations) per second. The integrated storage system will have more than 6 petabytes of storage with an I/O bandwidth of 140 gigabytes per second. The system is named after U.S. inventor and businessman Thomas Alva Edision. Read more.
ESnet, Infinera Demonstrate Prototype Switch for Dynamically Controlling Bandwidth Services
DOE’s ESnet and Infinera, a manufacturer of high capacity optical transmission equipment, demonstrated a prototype Software Defined Network (SDN) Open Transport Switch (OTS) capable of dynamically controlling bandwidth services. The demonstration marked the first time SDN was used to configure the network at the optical layer and was done using an extension to the OpenFlow protocol.
SDN is an emerging field in which software is used to create simple and powerful abstractions that allow application-driven configuration and control of the various layers of the network. ESnet is researching innovative ways to apply SDN value propositions to the wide-area network in order to improve the end-to-end experience of data-intensive science collaborations and applications.
The proof-of-concept demonstration, which used ESnet’s Long Island Metropolitan Area Network (LIMAN) testbed, tested a prototype of the OTS, allowing ESnet’s optical transport network to be configured by an SDN controller via the OpenFlow protocol. ESnet enhanced its SDN controller, integrated it with OSCARS, and demonstrated on-demand bandwidth Ethernet services including bandwidth elasticity for data-intensive science experiments at Brookhaven National Laboratory on their LIMAN network, spanning from Manhattan, NY, to Upton, NY.
ESnet’s OSCARS Software Helps Pacific Wave Peering Facility Users Make Stronger Connections
The TechZone 360 news site reports that Pacific Wave, a network peering facility connecting the West Coast, is using the On-Demand Secure Circuits and Advanced Reservation System (OSCARS) developed by ESnet to dynamically provision circuits between users connecting via the international peering facility. Pacific Wave is a joint project of the Corporation for Education and Network Initiatives in California (CENIC) and the Pacific Northwest Gigapop (PNWGP). OSCARS allows users to set up end-to-end network connections and reserve the necessary bandwidth for specified times. Read more.
ESnet5 Deployment Team Named as “Fierce” Innovators
The ESnet5 Deployment Team, charged with rolling out the 100 gigabit per second national network that entered production in November, has been named to the first annual "Fierce 15" list of the top federal employees and teams who have done particularly innovative things. The list was announced on November 27 by Fierce Government, a newsletter covering the U.S. government.
The list was compiled by Fierce Government staff who conducted interviews with government officials and industry to identify some of the most innovative projects and forward-thinking people working in government. From that pool, editors collaboratively selected this year’s Fierce 15. Read more.
CS Researchers to Present Findings at AGU Meeting in San Francisco
Michael Wehner, a climate researcher in Berkeley Lab’s Computational Research Division, will give an invited talk during the opening session of the American Geophysical Union’s fall 2012 meeting, which is being held December 3–7 in San Francisco. In his talk on “CMIP5 Projections of mean and extreme temperature and precipitation changes over the U.S.,” he will discuss how differences between the CMIP3 (Coupled Model Intercomparison Project Phase 3) and CMIP5 (Phase 5) projections of future temperature and precipitation changes are largely a result of differences in anthropogenic forcing rather than in the models themselves. A discussion of the similarities of the projected changes and their uncertainties will reveal that confidence in projections is increased by CMIP5.
Wehner is also a contributor to an invited talk by collaborator Ben Santer of Lawrence Livermore National Laboratory, “Identifying human influences on atmospheric temperature: Are results robust to uncertainties?”; a research paper, “The U.S. national climate assessment: The science of climate change”; and a research poster, “First contributions to the climate of the 20th Century Detection and Attribution Project.”
Other Computing Sciences researchers contributing to the AGU meeting include:
- Deb Agarwal (co-author): The Evolving AmeriFlux Network
- Alexandre Chorin (co-author): Numerical experiments with an implicit particle filter for the shallow water equations (poster)
- Val Hendrix, Keith Jackson, Lavanya Ramakrishnan, Keith Beattie, Christine Morin, David Skinner, Catharine van Ingen, Deb Agarwal, and others: Community Access to MODIS Satellite Reprojection and Reduction Pipeline and Data Sets (poster)
- Matthias Morzfeld, Ethan Atkins, and Alexandre Chorin: Implicit particle methods and their connection to weak and strong 4D-Var (poster)
- Arthur Wiedmer, Deb Agarwal, and others: Data-driven modeling of radionuclide inventory at the Savannah River Site F-area seepage basins and implications for long-term behavior predictions
Inder Monga Addresses International Network Architects Workshop
Inder Monga, ESnet’s Chief Technologist, gave a presentation on software-defined networking at TERENA’s first Network Architects Workshop, held Nov. 21–22 in Copenhagen. Software-defined networking, or SDN, is an emerging field in which software is used to automatically configure and control the various layers of the network to improve data transfers. ESnet has been conducting localized tests to demonstrate the value of SDN for production networking. Eric Pouyoul and Chin Guok co-authored the presentation.
The workshop, which was attended by 40 participants from 20 different organizations, provided a forum at which network designers and deployers shared their experiences and recommendations in building next-generation research and education networks. TERENA, which sponsored the workshop, is the Trans-European Research and Education Networking Association. The workshop agenda and slides can be found here.
Computerworld: Exascale Unlikely before 2020 due to Budget Woes
Computerworld magazine reports that the U.S. Department of Energy, which builds the world’s largest supercomputers, is now targeting 2020 to 2022 for an exascale system, two to four years later than earlier expectations. The new timeframe assumes that Congress will fund the project in the fiscal 2014 budget. Despite a belief among scientists that exascale systems can help deliver breakthrough scientific breakthroughs, improve U.S. competitiveness and deepen the understanding of problems like climate change, the development effort has so far received limited funding—nowhere near the billions of dollars likely needed. Read more.
Lab Women Sought as Mentors for High School Girls in Developing Mobile Apps
The Technovation Challenge, a 12-week program in which teams of high school and middle school girls compete to develop cool mobile apps, is again seeking mentors. The goal is to inspire girls to see themselves not just as users of technology, but as inventors, designers, builders and entrepreneurs. In the program, students work in teams to build apps, write business plans, and pitch their ideas to VCs for a chance to win $10,000 in funding. The challenge this year is to create an app that solves a problem in their local community. Go here or email AnnaLise Hoopes to learn more.
In spring 2012, Berkeley Lab hosted more than 60 girls in 11 teams from Berkeley and Albany high schools through the program and more than 20 women at the Lab volunteered as mentors. One team from Albany High went on to win both the regional and national competition with their app that combined social networking with preparation for college-level advanced placement tests.
Computing Sciences Holiday Party December 11
The Computing Sciences Winter Holiday Party will take place at the Toll Room, Alumni House, UC Berkeley campus, on Tuesday, December 11, from 4:00 to 8:00 pm. Refreshments will be served. Spouses and partners are welcome, but no children under 21.
For directions and parking go here.
Bay Area Scientific Computing Day Is December 12
The Institute for Computational and Mathematical Engineering at Stanford University is hosting this year’s Bay Area Scientific Computing Day (BASCD) on Wednesday, December 12, 2012. The event will be held in the Kavli Auditorium at the SLAC National Accelerator Laboratory.
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. Almost all speakers are junior researchers from various institutions in the San Francisco Bay Area, including graduate students and postdocs.
Two researchers from the Computational Research Division will be speaking at the event: Lin Lin will present “Fast algorithms for electronic structure analysis,” and Matthias Morzfeld will discuss “Implicit sampling for data assimilation and online filtering.”
If you are interested in attending, go here for additional information and registration.
This Week’s Computing Sciences Seminars
Dissertation Talk: Lightweight Specifications for Parallel Correctness
Wednesday, December 5, 11:00 am–12:00 pm, 373 Soda Hall, UC Berkeley
Jacob Burnim, UC Berkeley EECS
The spread of multicore processors and the end of rapidly growing single-core performance is increasing the need for programmers to write parallel software. Yet writing correct parallel programs with explicit multithreading remains a difﬁcult undertaking. Though many tools exist to help test, debug, and verify parallel programs, such tools are often hindered by a lack of any specification from the programmer as to the intended, correct parallel behavior of his or her software. In this talk, I will give an overview of our work on lightweight specifications for the parallelism correctness of multithreaded software. By focusing on just the correctness of a program’s use of parallelism, we have developed several novel, lightweight specifications for parallel software: semantic determinism, semantic atomicity, and nondeterministic sequential specifications for parallelism correctness. These specifications are not only simple for programmers to write — they enable us to much more effectively test, debug, and verify parallel software, independent of the software’s complex and sequential functional correctness.
Making Proof-Based Verified Computation Almost Practical
Wednesday, December 5, 12:00–1:00 pm, 310 Soda Hall, UC Berkeley
Michael Walfish, University of Texas, Austin
How can a machine specify a computation to another one and then, without executing the computation, check that the other machine carried it out correctly? And how can this be done without assumptions about the performer (replication, trusted hardware, etc.) or restrictions on the computation? This is the problem of verified computation, and it is motivated by the cloud and other third-party computing models. It has long been known that (1) this problem can be solved in theory using probabilistically checkable proofs (PCPs) coupled with cryptographic tools, and (2) the performance of these solutions is wildly impractical (trillions of CPU-years or more to verify simple computations).
I will describe a project at UT Austin that challenges the second piece of this folklore. We have developed an interactive protocol that begins with an efficient argument [IKO CCC ‘07] and incorporates new theoretical work to improve performance by 20+ orders of magnitude. In addition, we have broadened the computational model from Boolean circuits to a general-purpose model. We have fully implemented the system, accelerated it with GPUs, and developed a compiler that transforms computations expressed in a high-level language to executables that implement the protocol entities.
The resulting system, while not quite ready for the big leagues, is close enough to practicality to suggest that in the near future, PCPs could be a real tool for building actual systems.
Link of the Week: Science Faculty’s Subtle Gender Biases Favor Male Students
In the scientific disciplines in academia, there is a persistent disparity between the number of women receiving PhDs and those hired as junior faculty. Does gender bias contribute to this disparity? Experimental results published in the Proceedings of the National Academy of Sciences suggest that the answer is yes. Here is the abstract of the paper “Science faculty’s subtle gender biases favor male students” by Corinne A. Moss-Racusin et al.:
Despite efforts to recruit and retain more women, a stark gender disparity persists within academic science. Abundant research has demonstrated gender bias in many demographic groups, but has yet to experimentally investigate whether science faculty exhibit a bias against female students that could contribute to the gender disparity in academic science. In a randomized double-blind study (n = 127), science faculty from research-intensive universities rated the application materials of a student—who was randomly assigned either a male or female name—for a laboratory manager position. Faculty participants rated the male applicant as significantly more competent and hireable than the (identical) female applicant. These participants also selected a higher starting salary and offered more career mentoring to the male applicant. The gender of the faculty participants did not affect responses, such that female and male faculty were equally likely to exhibit bias against the female student. Mediation analyses indicated that the female student was less likely to be hired because she was viewed as less competent. We also assessed faculty participants’ preexisting subtle bias against women using a standard instrument and found that preexisting subtle bias against women played a moderating role, such that subtle bias against women was associated with less support for the female student, but was unrelated to reactions to the male student. These results suggest that interventions addressing faculty gender bias might advance the goal of increasing the participation of women in science.
One mitigation measure suggested by Pauline Gagnon of CERN, who circulated this study among women in the ATLAS experiment, would be to adopt a gender-blinding selection approach, where the gender would be revealed only after the short-listing process. This would ensure that the selection committee members would not unintentionally discriminate against female applicants at the very beginning of the selection process.
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.