ASCR Monthly Computing News Report - October 2010




Berkeley Lab’s Poznanski Looking at Alternative Yardstick to Measure the Universe
Astronomers have long relied on stellar explosions called Type Ia supernovas to measure the scale of the cosmos. A second class of supernovas may now be put to the same use, providing an independent check on measurements that were first used more than a decade ago to discover the accelerating expansion of the Universe. A growing number of researchers are working on the idea that some Type II supernovas — which are caused by the gravitational collapse of giant stars with iron cores — may have a role as gauges of cosmic distance. The method could be put to use with next-generation sky surveys — including the Dark Energy Survey due to start at Cerro Tololo in Chile in late 2011, and the Large Synoptic Survey Telescope, still in the development phase, at Cerro Pachón, also in Chile. Both are expected to find tens of thousands of supernovas a year.
“We’re at the stage where it would be stupid to ignore alternative methods to Type Ia,” says Dovi Poznanski, an astrophysicist at Lawrence Berkeley National Laboratory’s (Berkeley Lab’s) Computational Cosmology Center and the University of California, Berkeley, who has re-analyzed results that he says show the promise of the new cosmic measuring sticks. His most recent findings were published on 1 October in the Astrophysical Journal. In 2009, Poznanski demonstrated that a sample of 17 Type II supernovas could be used to predict the distances from Earth of their host galaxies with not much more than the error of estimates using Type Ia explosions.
Read the article in Nature News
Jaguar Helps Researchers Unravel the Magnetic Fields of Supernovas
A team led by Oak Ridge National Laboratory’s (ORNL’s) Tony Mezzacappa is getting closer to explaining the origins of core collapse supernova (CCSN) explosions with the help of the Jaguar supercomputer at the Oak Ridge Leadership Computing Facility (OLCF). Essentially, said Eirik Endeve, lead author of the team’s latest paper, researchers want to know how CCSN magnetic fields are created and how they impact the explosions of these massive stars. A recent suite of simulations allowed the team to address some of the most fundamental questions surrounding the magnetic fields of CCSNs. The team simulated a supernova progenitor, or a star in its pre-supernova phase.
New revelations in the team’s study mean two things to astronomers: first, that any rotation that serves as a key driver behind a supernova’s magnetism is created via the spiral mode; and second, that not only can the spiral mode drive rotation, but it can also determine the strength of a pulsar’s magnetic fields. Another major finding of the team’s simulations is that shear flow from the standing accretion shock instability (SASI), or when counter-rotating layers of the star rub against one another during the SASI event, is highly susceptible to turbulence, which can also stretch and strengthen the progenitor’s magnetic fields, similar to the expansion of a spring.
These two findings taken together show that CCSN magnetic fields can be efficiently generated by a somewhat unexpected source: shear flow-induced turbulence roiling the inner core of the star. “We found that starting with a magnetic field similar to what we think is in a supernova progenitor, this turbulent mechanism is capable of magnifying the magnetic field to pulsar strengths,” said Endeve. The findings were published in the April 20 issue of The Astrophysical Journal.
Contact: Jayson Hines, hinesjb@ornl.gov
Assessing Future Hurricane Impacts at the ALCF
Researchers at the National Center for Atmospheric Research (NCAR) and the Argonne Leadership Computing Facility (ALCF) are using the Nested Regional Climate Model (NRCM) on the ALCF’s Blue Gene/P supercomputer to develop improved assessments of impacts from Atlantic hurricanes over the next several decades. The NRCM combines the widely used Weather Research and Forecasting (WRF) model and the Community Climate System Model (CCSM). Through this combination, NRCM is able to effectively utilize both the climate simulation capacity of CCSM and the hurricane capacity of WRF to provide improved predictions of future hurricane activity to major societal and industry groups.
By using the ALCF facility, the NRCM is able to conduct the simulations at a very high resolution (equivalent to today’s hurricane forecast models) so that critical details of hurricane formation, structure and intensity can be simulated. This will further improve the prediction of hurricane impacts on our critical offshore energy infrastructure.
The work is contributing to a DOE-industry collaboration under the Research Partnership to Secure Energy for America (RPSEA) program, with emphasis on changing characteristics of hurricanes in the Gulf of Mexico offshore oil and gas fields. Research to date has indicated an increase in hurricane intensity through 2050. But this is countered by the hurricanes being smaller and faster, so that the net damage potential is predicted to decrease slightly.
Contact: James Done, done@ucar.edu
Repast SC++ Provides a Platform for Large-Scale Agent-Based Modeling
In the last decade, agent-based modeling and simulation (ABMS) has been successfully applied to a variety of domains, demonstrating the potential of this technique to advance science, engineering, and policy analysis. However, realizing the full potential of ABMS to find breakthrough research results requires far greater computing capability than is available through current ABMS tools. Repast Simphony for C++ (Repast SC++) addresses this need by developing a next-generation ABMS system explicitly focusing on larger-scale distributed computing platforms.
Leveraging years of experience in ABMS toolkit design and implementing the core Repast feature set, Repast SC++ is a useful and usable toolkit. It allows users to focus on model development and ignore the details of parallel programming. Written in portable C++, and using MPI and the Boost libraries, Repast SC++ achieves good performance at large scales. Simulation models can be written in C++, using the core simulation components directly, or in a Logo-style C++ that uses the typical Logo turtle, patch, link, and observer components.
The Repast SC++ toolkit has been implemented and tested on the IBM Blue Gene/P at the Argonne Leadership Computing Facility and was released as open source in October 2010. Repast SC++ is the subject of an invited chapter in the forthcoming book Large-Scale Computing Techniques for Complex System Simulations.
Contact: Michael North (north@anl.gov)


Medal of Science Winner Warren Washington has Ties to ASCR facilities
Warren Washington, who on October 19 was named by President Obama as a National Medal of Science winner, is a familiar name around the Oak Ridge Leadership Computing Facility (OLCF), Argonne Leadership Computing Facility and the National Energy Research Scientific Computing (NERSC) Center. The National Center for Atmospheric Research senior scientist and former chair of the National Science Board has collaborated with ASCR facilities since the 1990s.
According to Jim Hack of the OLCF and Climate Change Science Institute, Washington has been seminally involved in adapting global climate models to distributed-memory parallel computing environments. Washington has served as a principal investigator and advisor on LCF Climate allocations, including simulations cited in the fourth Intergovernmental Panel on Climate Change assessment report.
Read the White House Press Release
PNNL Researcher Critchlow Named Senior Member of IEEE
Terence Critchlow, a senior research scientist in Pacific Northwest National Laboratory’s (PNNL’s) Computational Sciences and Mathematics Division, has been elevated to the grade of Senior Member in IEEE, the world’s largest professional association dedicated to advancing technological innovation and excellence for the benefit of humanity. The organization has nearly 400,000 members, of which only 8% hold the Senior Member grade. Dr. Critchlow is the PNNL lead for the Scientific Data Management Center for Enabling Technologies funded by DOE OASCR.
Sandia’s Kolda Named Section Editor for SIAM J. Scientific Computing
Tamara G. Kolda of Sandia National Laboratories has been named Section Editor of the newly formed Software and High-Performance Computing (S&HPC) section of the SIAM Journal on Scientific Computing (SISC); she has previously served two three-year terms as an associate editor. The journal has recently revised its editorial policy and has a new organizational structure with three distinct sections. The newly formed S&HPC section will feature articles on the development of high quality computational software, high performance computing issues, novel architectures, data analysis, and visualization. For more information, visit SISC Redefined at the following link:
ORNL Staff Fill Key Roles at SC10 Conference
ORNL staff will fill a variety of leadership roles at the 23rd annual international Supercomputing conference (SC10) to be held November 13–19 in New Orleans. Thirty-two representatives from ORNL will take part in a variety of organizational and technical aspects of the conference, including 12 staff members either chairing or co-chairing various events, and five members working with SCinet, the advanced network that powers the exhibitions and demonstrations throughout the conference.
Becky Verastegui, ORNL Information Technology Services director, is SC10’s vice-chair, and Ricky Kendall, Scientific Computing Group leader, is chair of the technical program. Jim Hack, director of the National Center for Computational Sciences, will be co-chairing one of the conference’s main technological thrusts: climate simulation. Computational scientists Hai Ah Nam and Scott Klasky will chair the student cluster competition and the storage programs, respectively; and network services engineer Charles Fisher will be chairing the conference’s session on high performance computing (HPC) architecture. The tutorial program, designed to show conference attendees how to use some of the large-scale applications for supercomputers, will be chaired by senior HPC researcher Jeffery Kuehn. Doug Fuller, input/output systems computational scientist, will co-chair the poster session, at which more than 60 teams will present cutting-edge research throughout the course of the weekend.
Contact: Jayson Hines, hinesjb@ornl.gov
Berkeley Lab’s Agarwal, Aragon Share Expertise at 2010 Grace Hopper Conference
Deb Agarwal, head of the Advanced Computing for Science Department (ACS) at Berkeley Lab, and Cecilia Aragon, a researcher in ACS, each participated in several sessions at the 2010 Grace Hopper Celebration of Women in Computing Conference, held September 28–October 2 in Atlanta. Agarwal served as a mentor in the Ph.D. Forum, and was a panelist in discussions on “How do I become a leader in my field?” and “Enabling the Next Generation of Scientific Breakthroughs via Computer Science.” Aragon, who is also an associate professor at the University of Washington, was a presenter in the career mentoring session, speaking on “How do I become a researcher?” and participated in panel discussions on “Women of Color: Strategies for Excelling and Thriving,” “Mastering the Art of the Technical Interview” and “Speed Mentoring for Latinas in Computing.”
Contact: Jon Bashor, jbashor@lbl.gov
Berkeley Lab’s Kathy Yelick Delivers Keynote at ISC Cloud Conference in Frankfurt
Kathy Yelick, director of the NERSC Division and Associate Laboratory Director for Computing Sciences at Berkeley Lab, delivered a keynote address at the ISC Cloud ’10 conference held October 28–29 in Frankfurt, Germany. Yelick discussed “Science in the Clouds: The View from Berkeley.”
Building on the NERSC experience serving a broad scientific workload with over 3000 users, her talk examined some of the characteristics of scientific applications, including both data- and compute-intensive problems, and the suitability of the cloud computing model. She also discussed how scientific clouds should be configured from a hardware and software standpoint, and reported on some of the early successes with DOE’s Magellan cloud testbed for bioinformatics applications. Yelick also participated in a discussion entitled “Cloud Debate: Cloud or Not Cloud, That Is the Question.”
Contact: Jon Bashor, jbashor@lbl.gov


ESnet Announces Release of Version 3.2 of perfSONAR pS Performance Toolkit
The perfSONAR collaboration, which includes ESnet, Internet2 and others, has just announced the release of version 3.2 of the pS Performance Toolkit. perfSONAR is an infrastructure for network performance monitoring, making it easier to solve end-to-end performance problems on paths crossing several networks. It contains a set of services delivering performance measurements in a federated environment. The latest version of perfSONAR-PS is available at this link.
Contact: Wendy Tsabba, wtsabba@lbl.gov
Mike Helm Outlines Science Identity Federation at Authentication Technologies Meeting
Mike Helm of ESnet gave a presentation on the Science Identity Federation at the Symposium on Authentication Technologies for Research and Education, held October 4–7 at Texas Tech University in Lubbock, Texas. The Science Identity Federation will provide the basis for interoperable identity and identity-based services for DOE national laboratories and their collaborators: academia, other laboratories, and other organizations.
Contact: Wendy Tsabba, wtsabba@lbl.gov


Argonne, Berkeley “Collaborate Across Boundaries” at 2010 Grace Hopper Conference
Argonne National Laboratory and Lawrence Berkeley National Laboratory served as Bronze Sponsors of the 2010 Grace Hopper Celebration of Women in Computing Conference, held September 28–October 2 in Atlanta, Ga. This year’s theme was “Collaborating Across Boundaries.” Designed to bring the research and career interests of women in computing to the forefront, the annual Hopper meeting is the largest technical conference for women in computing, and results in collaborative proposals, networking and mentoring for junior women, and increased visibility for the contributions of women in computing. Conference presenters are leaders in their respective fields, representing industry, academia, and government. The 2010 conference drew 2,147 attendees, 960 students, 280 schools, 29 countries, and 630 speakers. For more information, visit the conference website the following link:
ALCF Early Science Program Workshop Offers a Peek at Next-Generation Blue Gene
The Argonne Leadership Computing Facility (ALCF) hosted an Early Science Program (ESP) Kick-Off Workshop on Oct.18–19 at Argonne National Laboratory. The by-invitation-only workshop represented the first gathering of key players in this exciting endeavor to build the future of computational science. The event drew 77 participants from 23 institutions—39 were representatives from 16 institutions for ESP projects. At the workshop, IBM staff provided a first look at the next-generation Blue Gene hardware and software, including compilers and messaging. ESP project teams presented overviews of their projects and plans for their first six months of development. Also presented were performance measurements, as well as efforts being undertaken in the next-generation Blue Gene Tools, Libraries, and Programming Models project. Hands-on help was available for those new to Blue Gene to facilitate the development efforts key to the ESP.
Contact: David Martin, dem@alcf.anl.gov, or Chel Heinzel (chel@alcf.anl.gov)
ITAPS SciDAC Team Conducts Short Course on Parallel Meshing
Karen Devine (Sandia National Laboratories) and Timothy Tautges (Argonne National Laboratory) presented a short course, “Parallel Mesh Management using Interoperable Tools,” at the 2010 International Meshing Roundtable on October 3. Devine and Tautges are members of the DOE SciDAC ITAPS (Interoperable Technologies for Advanced Petascale Simulations) team. Their presentation included a discussion of challenges arising in parallel mesh management, as well as demonstrated solutions. They also described the broad range of software for mesh management and modification developed by the ITAPS team, and highlighted applications successfully using the ITAPS tool suite.
Berkeley Lab Staff Contribute to IEEE VisWeek 2010 Conference
IEEE VisWeek 2010 was held October 24–29 in Salt Lake City, Utah, and members of Berkeley Lab’s Visualization Group and NERSC Analytics Team contributed to a tutorial and a paper. Hank Childs co-organized the tutorial on “Large Vector-Field Visualization: Theory and Practice.” Gunther Weber co-authored a paper on “Two-Stage Framework for a Topology-Based Projection and Visualization of Classified Document Collections.” Also, Inna Dubchak of Berkeley Lab’s Genomics Division joined a panel discussion of “Challenges in Visualizing Biological Data.”
Contact: Jon Bashor, jbashor@lbl.gov