# InTheLoop | 01.04.2010

January 4, 2010

## Computational Cosmology Center Becomes Independent Group

Several years ago, recognizing their decade-long collaboration in this field, the Physics and the Computational Research (CRD) divisions at LBNL formed the Computational Cosmology Center (C3). C3 is a focused collaboration of astrophysicists and computational scientists whose goals are to develop the tools, techniques, and technologies to meet the analysis challenges posed by present and future cosmological data sets. C3 has been a very effective collaboration, and the projects within C3 have grown significantly, as recent successes of the group related to projects such as Deep Sky, Palomar Transient Factory, and Planck have shown.

As astronomy is increasingly becoming a data rich science, there are significant new opportunities in the field of computational astrophysics and cosmology for C3 to explore. Therefore, effective today (January 4, 2010), C3’s status will be recognized as an independent group within the High Performance Computing Research Department in CRD.

The scientific leadership of C3 will continue to be in the hands of Julian Borrill and Peter Nugent. In addition, Peter Nugent has agreed to take on the responsibilities of a group lead for C3 within CRD. The following CRD staff scientists and visitors will transfer from the Scientific Computing Group to C3: Edward Baron, Julian Borrill, Chris Cantalupo, Theodore Kisner, Peter Nugent, Dovi Poznanski, Rajesh Sudarsan, Radek Stompor, and Rollin Thomas.

In his announcement of this change, Associate Laboratory Director Horst Simon said, “I would like to thank Esmond Ng who as group lead of the Scientific Computing Group has supported the C3 project for many years and helped to grow it into a large group that is now becoming independent. Please join me and welcome Peter Nugent as a new group lead, and wishing C3 the best of success in the future.”

## Kamesh Madduri Is First Winner of SIAG/SC Junior Scientist Prize

Kamesh Madduri, an Alvarez Fellow working in CRD’s Scientific Data Management Research Group, has been selected as the first winner of the Junior Scientist Prize established by the SIAM Activity Group on Supercomputing (SIAG/SC). This prize is awarded to an outstanding junior researcher in the field of algorithms research and development for parallel scientific and engineering computing, for distinguished contributions to the field in the three calendar years prior to the year of the award.

Madduri’s research interests include parallel graph algorithm design for complex network analysis, scientific data analysis query optimization using bitmap indexes, and performance tuning of particle-in-cell simulations on multicore architectures. He is one of the principal developers of SNAP (Small-World Network Analysis and Partitioning), an extensible parallel framework for exploratory analysis and partitioning of large-scale networks.

The prize will be awarded at the 2010 SIAM Parallel Processing Conference in Seattle, February 24–26. At the conference, Madduri will give a 15-minute presentation on his research.

## Hiring Guidelines for Computing Sciences Are Revised

The Hiring Package section of Computing Sciences’ Recruiting, Interviewing, and Hiring Guidelines has been revised with the addition of a new section on postdoctoral fellows. Hiring packages for scientists and CSEs should now include a copy of the interview schedule and a copy of the seminar announcement with title and abstract (if a seminar was given). Everyone who has a role in the hiring process is encouraged to review these guidelines.

Two NERSC users, William Dorland and Wim Leemans, are among the six winners of the 2009 E.O. Lawrence Award for their outstanding contributions in research and development supporting the Department of Energy and its missions. Each winner will receive a gold medal, a citation and $50,000, and will be honored at a ceremony in Washington, DC. William Dorland of the University of Maryland will be honored for his scientific leadership in the development of comprehensive computer simulations of plasma turbulence, and his specific predictions, insights, and improved understanding of turbulent transport in magnetically confined plasma experiments. Wim Leemans of Lawrence Berkeley National Laboratory will be honored for his breakthrough work in developing the laser plasma wakefield accelerator from concept to demonstration, and his scientific leadership exploring its promise and unprecedented possibilities ranging from hyperspectral light sources to high energy colliders. ## Berkeley Water Center Featured in “Socially Relevant Computing” Deb Agarwal’s research collaboration with Microsoft Research and the Berkeley Water Center is one of several projects featured in a brochure from Microsoft Research’s External Research Division. The brochure, titled “Socially Relevant Computing: Our Health, Our Knowledge, Our Planet,” is available online. A section titled “Making Sense of Data: Helping Scientists Understand Our Changing Environment” (pages 38–41) says: Environmental researchers at the Berkeley Water Center in California have teamed up with computer scientists from Microsoft and the Lawrence Berkeley National Laboratory to create powerful new databases and software collaboration tools that are helping scientists to understand our changing environment and develop more reliable models for resource management…. Agarwal says the goal of the data server project is to provide “game-changing tools” that enable environmental scientists to easily browse large volumes of data without having to understand the underlying computer science. ## Give a Humanist a Supercomputer … “… and you’ll be surprised what he or she can do with it,” according to an article in The Chronicle of Higher Education. The article reports on the results of the National Endowment for the Humanities and the Department of Energy’s Humanities High Performance Computing Program, which were on display at the Coalition for Networked Information membership meeting held in Washington, DC, in December. Recipients of Humanities HPC Program grants used NERSC computers to advance their research. ## NERSC to Co-Host Cray XT5 Workshop in February NERSC, along with the Oak Ridge Leadership Computing Facility (OLCF) and the National Institute for Computational Science (NICS) at the University of Tennessee and Oak Ridge National Laboratory, will present a joint Cray XT5 Workshop February 1–3, 2010, at Sutardja Dai Hall on the UC Berkeley campus. The workshop is designed to provide an in-depth introduction to using the world’s newest and largest Cray XT5 systems. Representatives and staff from NERSC, OLCF, NICS, Cray, and AMD will explain how to use XT5 systems productively. Katie Antypas and Richard Gerber from NERSC are among the presenters. The workshop is aimed at both new and intermediate users of the Cray XT5 who already have some high performance computing experience — knowledge of Linux, a programming language such as C/C++ or Fortran, and some exposure to parallel programming concepts with the Message Passing Interface (MPI). Hands-on sessions will use Cray XT systems at NERSC, OLCF, and NICS. ## HotPar ’10 Is Accepting Papers for June Workshop in Berkeley The Second USENIX Workshop on Hot Topics in Parallelism (HotPar ’10) will be held June 14–15, 2010 in Berkeley, CA. (David Patterson of UC Berkeley and CRD is Program Co-Chair.) HotPar ’10 will bring together researchers and practitioners doing innovative work in the area of parallel computing. To ensure a productive workshop environment, attendance will be limited to 75 participants. Multicore processors are the pervasive computing platform of the future. This trend is driven by limits on energy consumption in computer systems and the poor energy performance of conventional microprocessors. Parallel architectures can potentially mitigate these problems, but this new computer architecture will only be successful if languages, systems, and applications can take advantage of parallel hardware. Navigating this change will require new concurrency-friendly programming paradigms, new methods of application design, new structures for system software, and new models of interaction between applications, compilers, operating systems, and hardware. HotPar requests submissions of position papers that propose new directions for research of products in these areas, advocate non-traditional approaches to the problems engendered by parallelism, or potentially generate controversy and discussion. Submissions are due by January 24, 2010. ## This Week’s Computing Sciences Seminars An Asymptotic Preserving Hybrid Godunov Method for Radiation Hydrodynamics Monday, January 4, 9:30–10:30 am, 50B-4205 Michael Sekora, Princeton University From a mathematical perspective, radiation hydrodynamics can be thought of as a system of hyperbolic balance laws with dual multiscale behavior (multiscale behavior in the hyperbolic wave speeds as well as multiscale behavior in the source term relaxation). However, these dual behaviors are quite different and cause breakdowns in monotonicity and stability, respectively, while influencing the temporal resolution of the problem. These considerations are taken into account when designing a robust algorithm. This talk presents a hybrid Godunov method for one-dimensional radiation hydrodynamics that is asymptotically preserving and uniformly well behaved from the photon free streaming (hyperbolic) limit through the weak equilibrium diffusion (parabolic) limit and to the strong equilibrium diffusion (hyperbolic) limit. The method incorporates a backward Euler upwinding scheme for the radiation energy density$E_{r}$and flux$F_{r}$as well as a modified Godunov scheme for the material density$\rho$, momentum$m$, and energy density$E\$.

The backward Euler scheme is first-order accurate and uses an implicit HLLE flux function to temporally advance the radiation components according to the material flow scale. The modified Godunov scheme is second-order accurate and directly couples stiff source term effects to the hyperbolic structure of the system of conservation laws. This Godunov technique is composed of a predictor step that is based on Duhamel’s principle and a corrector step that is based on Picard iteration. The Godunov scheme is explicit on the material flow scale but is unsplit and fully couples matter and radiation without invoking a diffusion-type approximation for radiation hydrodynamics. This technique derives from earlier work by Miniati & Colella 2007. Numerical tests demonstrate that the method is stable, robust, and accurate across various parameter regimes.

## Link of the Week: Babbage’s Difference Engine

Charles Babbage, the man whom many consider to be the father of modern computing, never got to complete any of his life’s work. The Victorian gentleman was a brilliant mathematician, but he wasn’t very good at politics and fundraising, so he never got the financial backing to finish any of his elaborate machine designs. For decades, even his fans weren’t certain whether his computing machines would have worked.

But Doron Swade, a former curator at the Science Museum in London, has proven that Babbage wasn’t just an eccentric dreamer. Using nothing but materials that would have been available to Babbage in the 1840s, Swade and a group of engineers successfully built Babbage’s Difference Engine, and a version is now on display at the Computer History Museum in Mountain View, Calif. Read more from NPR.

##### About Berkeley Lab Computing Sciences

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 5,500 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.