InTheLoop | 11.10.2008
The weekly newsletter for Berkeley Lab Computing Sciences
November 10, 2008
LBNL Researchers Contribute Expertise in All Aspects of SC08 Conference
LBNL researchers are making significant contributions to the SC08 Conference Technical Program, contributing four technical papers and one research poster, organizing two workshops, participating in two panel discussions, and hosting or co-hosting six birds-of-a-feather sessions (BoFs). SC08, the international conference on high performance computing, networking, storage and analysis, will be held Nov. 15–21 in Austin, Texas.
Additionally, UC Berkeley Prof. David Patterson, who has a joint appointment in LBNL’s Future Technologies Group, is one of four invited speakers; and Dale Sartor of the Environmental Energy Technologies Division will give a Masterworks presentation on energy-efficient computing.
As mentioned in the Sept. 15 InTheLoop, the developers of the LS3DF electronic structure code, led by Lin-Wang Wang, are candidates for the ACM Gordon Bell Prize, which will be awarded at SC08. For a story about their work, go to InTheLoop.
CRD Safety Stand-Down Meetings Today and Wednesday
As part of the Lab’s overall effort to improve employee safety and safety awareness, and to prepare for the HSS review, CRD will be holding three safety stand-downs next week. Every division member is required to attend one of the sessions, which are being organized by group or department. If you cannot attend the meeting scheduled for your group, you must attend one of the other sessions. CRD administrative staff should attend with their designated groups. If you are unable to attend any of the three sessions, please inform your group leader. Here is the schedule:
Monday, Nov. 10, 1-3 p.m., Room 50A-5132
- ESnet Department
Wednesday, Nov. 12, 10 a.m.-noon, Bldg. 50 Auditorium
Groups in Bldg. 50 A:
- Applied Numerical Algorithms Group
- Center for Computational Sciences & Engineering
- Future Technologies Group
- Mathematics Group
- Biological Management and Technology Center
Wednesday, Nov. 12, 1-3 p.m., Bldg. 50 Auditorium
Groups in Bldg. 50B and 50F:
- Scientific Data Management Group
- Advanced Computing for Science Department
- Scientific Data Management Group
- Scientific Computing Group
- Visualization Group
Betsy MacGowan Joins CS as Full-Time Safety Coordinator
Betsy MacGowan, who has been the EH&S Liaison to Computing Sciences for the past three years, now joins the organization as Safety Coordinator. MacGowan, who started her new position on Nov. 3, will take over the safety coordination responsibilities of John Hutchings, allowing him to fully focus on his roles as CS Facilities Manager and Bldg. 50 Building Manager. Now as a full-time safety coordinator, she said she will focus on raising employee awareness of ergonomic and personal safety issues in the workplace.
MacGowan first came to the Lab in 2001 as a consultant working on a seismic retrofitting of Bldg. 77. She said she immediately took to the place and began looking for a full-time position here. She landed a job as a contractor in the Industrial Hygiene Group in the EH&S Division, monitoring asbestos and lead abatement work. After a year, she was hired into the group as a LBNL employee.
She earned her bachelor’s degree in environmental science at the University of Virginia and worked as an asbestos abatement consultant prior to joining the Lab. MacGowan is a certified industrial hygienist as well as a certified safety professional. She can be contacted at EEMacgowan@lbl.gov.
“One of Betsy’s first tasks will be preparing Computing Sciences for the upcoming safety audit by DOE,” said ALD Horst Simon. “Please welcome Betsy to CS and give her your full support in this important task.”
The new EH&S Liaison for Computing Sciences will be Scott Robinson. Scott is new to LBNL, but he has over 25 years of EH&S experience. He is also responsible for the Lab’s Indoor Air Quality and Ventilation Programs, which pair well with this assignment. He can be contacted at STRobinson@lbl.gov.
Help Identify Game-Changing Advances from Computing Research
The Computing Community Consortium is asking the research community to help identify “game-changing advances from computing research conducted in the past 20 years.” Initial candidates for the list include the Internet and World Wide Web, searching unstructured information, cluster computing, and the transformation of science via computation.
To critique these choices or suggest your own, go to The Computing Computing Consortium Blog.
This Week’s Seminar Schedule
Wednesday, Nov. 12, 11:10 am-noon, 380 Soda Hall, UCB
Simulation of Lean Premixed Turbulent Combustion
John Bell, LBNL/CCSE
Combustion is one of our oldest and most important technologies. It continues to provide most of the energy required for transportation and power generation. Recent concerns over U.S. dependence on imports of foreign oil, coupled with pollution and greenhouse gas emission issues, have generated significant interest in developing new fuel-flexible combustion systems that can employ alternative fuels such as hydrogen, ethanol or syngas. Effective utilization of these fuels requires combustion devices that can operate cleanly and efficiently over a broad range of fuels and fuel mixtures. Lean-premixed systems have the potential for meeting these requirements; they operate at high efficiency and have low NOx emissions due to reduced burnt gas temperatures. Although traditional scientific approaches based on theory and laboratory experiment have played essential roles in developing our current understanding of premixed combustion, they are unable to meet the challenges of designing fuel-flexible lean, premixed combustion devices. Computation, with its ability to deal with complexity and its unlimited access to data, has the potential for addressing these challenges.
Realizing this potential requires the ability to perform high fidelity simulations of turbulent lean premixed flames under realistic conditions. Estimates of the computational cost of such simulations suggest that naive simulation approaches would be prohibitively expensive. In this talk, we examine the specialized mathematical structure of combustion problems and discuss approaches to simulation that exploit this structure. Using these ideas we can reduce computational cost by three to four orders of magnitude, making it possible to perform high-fidelity simulations of realistic laboratory flames. We will illustrate this methodology by considering several configurations to illustrate the methodology and discuss how this type of simulation is changing the way researchers study combustion.
Wednesday, Nov. 12, noon–1:00 pm, 290 Hearst Memorial Mining Bldg., UCB
CITRIS Research Exchange
Pergamum: Energy Efficient, Reliable, Disk-Based Archival Storage
Ethan Miller, Professor of Computer Science, UC Santa Cruz
Live webcast: mms://media.citris.berkeley.edu/webcast
As the world moves to digital storage for archival purposes, there is an increasing demand for reliable, low-power, cost-effective, easy-to-maintain storage that can still provide adequate performance for information retrieval and auditing purposes. Unfortunately, no current digital archival system — tape, disk, or optical disk — adequately fulfills all of these requirements. To address this challenge, we developed Pergamum, which stores data in a network of “bricks,” each of which contains a disk, low-power CPU, and flash memory. In normal operation, only a small fraction of the disks in a Pergamum system are spun up, dramatically reducing power consumption. Since the other components use very little power, a Pergamum system with 5% of the disks in use consumes less than 2 watts per terabyte of stored data.
Pergamum is also designed to be very highly reliable. Pergamum stores additional redundant information on each brick, allowing the brick to repair itself in many cases, and takes advantage of the bricks’ network connections and low-powered flash memory to constantly verify storage integrity between bricks without the need to constantly spin up disks, ensuring data preservation with low power requirements. If an error is found, Pergamum can rebuild the lost data with low peak power consumption, avoiding the need for large-scale power and cooling infrastructure. Pergamum is also designed to evolve over time: it is based around standard IP-based networking protocols, and can accommodate bricks using any storage technology, as long as the bricks “speak” the standard protocols. By using these techniques, Pergamum can provide archival storage at cost and power consumption comparable to tape with far better performance, reliability, and evolvability than a tape-based system. Current research on Pergamum is exploring issues with integrating new devices and retiring old devices, failure detection in a system with hundreds of thousands of devices, and search across exabyte-scale storage in a power-constrained environment.
Wednesday, Nov. 12, 4-5 pm, 540 Cory Hall, UCB
The Nimrod/K Director for the Kepler Workflow Environment
Colin Enticott, Monash University, Australia
A challenge for Grid computing is the difficulty in developing software that is parallel, distributed and highly dynamic. In this presentation we will discuss the challenges of providing a new Grid system by combining the workflow system Kepler with a large scale Grid middleware application called Nimrod. The Kepler project provides a workflow environment with various tools to access Grid services. The Nimrod toolkit offers Grid services such as multi-resource meta-scheduling, data movement and parameter search and analysis tools. We will discuss how grid applications and their data files are represented in Nimrod/K and how we have extended the Kepler environment to support threading. Combining these two Grid applications offers a powerful environment for performing large scale Grid base workflows.
Thursday, Nov. 13, 2:00–3:00 pm, 290 Hearst Memorial Mining Bldg., UCB
Nokia Distinguished Lecture: A Theory of Robustness for Cyber-Physical Systems
George Pappas, Electrical and Systems Engineering Dept., University of Pennsylvania
Live webcast: mms://media.citris.berkeley.edu/webcast
One of the great challenges in cyber-physical systems is to define appropriate measures of system robustness. How should we define the robust execution of digital programs by physical systems that will be subject to uncertainty and noise? By reversing roles, how robust is the digital implementation of physical controllers and sensors in distributed computing platforms? Can we define appropriate measures of robustness across the cyber and physical world, in a manner that leads to efficient algorithms for overall system verification and design?
In this talk, I will present an overview of our efforts towards addressing some of these challenges. Thinking of temporal logics as basic programming languages for physical systems, such as robots, we define robust, multi-valued semantics for temporal logic formulas, which capture not only the usual Boolean satisfiability, but also how robustly the physical system satisfies the digital specification. Based on this quantitative notion and using our recently developed notion of approximate bisimulation functions, we develop a simulation-based verification algorithm for determining the robustness of the overall system. The interesting and promising feature of our approach is that the more robust the system is with respect to the temporal logic specification, the less is the number of simulations that are required in order to verify the system. We then consider the problem of quantifying the performance gap between model-based controller design and platform-based implementation for modern embedded control systems. We first show that the performance gap between the model-level semantics of proportional-integral-derivative (PID) controllers and their implementation-level semantics can be rigorously quantified if the controller implementation is executed on a predictable time-triggered architecture. Explicitly computing the impact of the implementation on overall system performance allows us to compare and partially order different implementations with various scheduling or timing characteristics.
Link of the Week: The Shape of Music
Roughly 2,500 years ago, Pythagoras observed that objects produced harmonious sounds while vibrating at frequencies in simple whole-number ratios. Centuries later, Gottfried Leibniz described music as the “unknowing exercise of our mathematical faculties.”
Recently Dmitri Tymoczko and other theorists have shown that the description of the melodic and harmonic structure of Western music requires the geometry and topology of what mathematicians call “quotient spaces” or “orbifolds.” These exotic spaces contain singularities — points that are analogous to the black holes of Einstein’s general relativity — that can be described using only very recent mathematics. “One can argue that Romantic composers such as Chopin had an intuitive feel for non-Euclidean higher-dimensional spaces that exceeded the explicit understanding of their mathematical contemporaries,” Tymoczko writes in Seed magazine. The article can be read at SEED.
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.