InTheLoop | 11.22.2010
November 22, 2010
Magellan Wins “Best Use of HPC in the Cloud” Award
The U.S. Department of Energy’s cloud computing testbed project known as “Magellan” has been recognized in the annual HPCwire Readers’ and Editors’ Choice Awards. Representatives of Argonne and Lawrence Berkeley national laboratories, where Magellan testbeds are located, were presented the award at SC10, the international conference for high performance computing, networking, storage and analysis, in New Orleans.
The Magellan project, which is exploring the suitability of cloud computing to help meet the computational science requirements of DOE researchers, was honored with the Readers’ Choice Award for “Best Use of HPC in the Cloud.” Read more.
ESnet and Collaborators Achieve 100 Gbps Data Flow to SC10
A team including NASA, NOAA, ESnet, the Dutch Research Consortium, US LHCNet, and CANARIE announced last week that they had transmitted 100 Gbps of scientific data between Chicago and SC10 in New Orleans. It was “the first time for a supercomputing conference that a single wavelength over the WAN achieved 100 Gbps,” according to ESnet’s Brian Tierney. Read more. For more blog entries about ESnet at SC10, go here.
Kesheng John Wu Is Named ACM Distinguished Scientist
ACM (the Association for Computing Machinery) has named 47 of its members as Distinguished Members in recognition of their individual contributions to practical and theoretical aspects of computing that drive innovation and sustain economic competitiveness. Kesheng John Wu of the Scientific Data Management Research Group in CRD is one of 41 ACM members named 2010 Distinguished Scientists. Go here for the complete list.
Hasenkamp Wins Third Place in ACM Student Research Poster Competition
Daren Hasenkamp, an undergraduate intern from UC Berkeley, has won third place in the SC10 ACM Student Research Competition for his poster on “Finding Tropical Cyclones on Clouds.” As one of the top finishers, he will be invited to participate in the ACM Student Research Competition Grand Finals. A full paper on his work with Alex Sim, Michael Wehner, and Kesheng John Wu has been accepted at CloudCom 2010. Here is the poster abstract:
Extensive computing power has been used to tackle issues such as climate changes, fusion energy, and other pressing scientific challenges. In this work, we bring the power of cloud computing to bear on the task of analyzing trends of tropical cyclones in climate simulation data. The cloud computing platform is attractive here because it can provide an environment familiar to climatologists and their analysis tools. We created virtual machines (VMs) and ran them on the Magellan Scientific Cloud at Argonne National Laboratory. Our VM communicates with instances of itself to split up and analyze large datasets in parallel. In a preliminary test, we used this virtual climate analysis platform to analyze ~500 GB of climate data. Using 34 VMs, the total analysis time was reduced by a factor of ~40 from the traditional analysis method. The main advantages of our method are that the level of parallelism is easily configurable, and software dependency resolution is simple.
Job Open for Assistant OSF Facility Manager
The NERSC Division has an opening for an Assistant Facility Manager at the Oakland Scientific Facility, providing support to the NERSC Division for activities which include operations, site preparation, maintenance, planning, and assisting with project management of the NERSC supercomputer room and associated building support infrastructure consisting of medium and low voltage electrical distribution and chilled water mechanical systems. The Assistant Facility Manager will also support the staff office areas including employee and equipment relocations, LBNL asset tracking, and assisting in the resolution of Environment, Health and Safety (EH&S) issues. This position will act as a backup for the OSF Facility Manager as needed. See job details.
Applications Open for Access to the Yahoo! Cloud Cluster
Originally launched in November 2007, Yahoo!’s M45 cloud cluster program is providing seven universities, including UC Berkeley, the opportunity to conduct research otherwise impossible without the power and speed of a supercomputing resource, which consists of approximately 4,000 processors.
The UC Berkeley CITRIS program research projects facilitated on Yahoo!’s M45 supercomputer include analyzing social networks, studying population genetics, testing new architectures for collecting traffic data from GPS-equipped mobile phones and estimating traffic conditions in real-time, analyzing climate-change satellite data, prototyping new scientific applications, and improving cluster scheduling and reliability.
To apply for access to this cloud resource, please contact Dr. Masoud Nikravesh, CITRIS Director for Computational Science and Engineering (i4Science).
Institute for Computing in Science Is Accepting Proposals
The Institute for Computing in Science (ICiS) is accepting program proposals for its 2011 summer and fall programs. The 2011 summer program of eight one-week workshops will be held at The Canyons Resort in Park City, Utah. The fall 2011 program will be held at the Theory and Computing Science (TCS) building at Argonne National Laboratory.
The objective of these programs is the development of an interdisciplinary approach to solving fundamental problems in science and engineering through advances in computational methods, algorithms and software, scientific computing technology and architecture, and the innovative application of large-scale data analysis.
This Week’s Computing Sciences Seminars
Identifying Human Influence on Extreme Precipitation
Monday, Nov. 22, 11:30 am–12:30 pm, 50F-1647
Seung-Ki Min, Environment Canada
Weather and climate extreme events can have devastating effects on human society and the environment, and understanding past changes in such events is critical for more reliable future projections. Heavy precipitation events have become more intense over a large part of the Northern Hemisphere land area. It has been suggested that human-induced increase in greenhouse gases may be partly responsible for the observed intensification of extreme precipitation, as expected from the increased water holding capacity of atmosphere with global warming. However, because of the limited availability of daily observations, most previous studies have examined only the potential detectability of changes in extreme precipitation through model-model comparisons. Here we compare observed and multi-model simulated extreme precipitation using an optimal fingerprinting technique, and provide the first evidence that human activity has contributed to the observed intensification of heavy precipitation events during the latter half of the 20th century. Models, however, tend to underestimate the observed increase in heavy precipitation trend, indicating that model projected future changes in extreme precipitation and their impacts may also be underestimated.
Bypass Based IP Offloading Using Dynamic Circuit Services
Monday, Nov. 22, 1:00–2:00 pm, 50B-2222
Mohit Chamania, Technische Universität Carolo-Wilhelmina zu Braunschweig, Germany
Research efforts towards optimizing the performance of IP networks have been focused on increasing the utilization of the network while minimizing additional resources used. Many such paradigms developed under the umbrella of Traffic and Network Engineering are either cost inefficient or can lead to frequent and significant routing changes in the network when used to tackle short-lived traffic churns, and are therefore not commonly used by network providers. In this work, we present a new IP traffic offloading paradigm which is targeted towards improving network utilization by offloading traffic onto dynamic circuits while ensuring that the IP routing protocol remains unchanged. The approach intrinsically ensures that reconfiguration effort in the network is low and is more cost efficient than a simple capacity upgrade of IP links. We show comparative performance of the bypass-based offloading paradigm using a few case studies, and also talk about implementation aspects currently being studied at IDA.
Related to our work on bypass and hybrid networking, this talk will also present an overview of the European Project ONE. The project ONE, which we are coordinating, is a new management adapter to enable cross-layer operations for coordination of IP and underlying transport networks via interactions between their management planes.
HPC at LRZ and in Europe
Tuesday, Nov. 23, 9:00–10:00 am, OSF 236
Prof. Dr. Dieter Kranzlmüller, Leibniz Rechenzentrum (LRZ), Munich, Germany
Par Lab Seminar: Transactional Language Constructs for C++
Tuesday, Nov. 23, 11:00 am–12:30 pm, 438 Soda Hall (Wozniak Lounge), UC Berkeley
Tatiana Shpeisman, Intel
Transactional memory (TM) promises to simplify concurrent programming by providing a concurrency control abstraction that avoid pitfalls associated with lock-based synchronization. Over the last decade, much progress has been made in hardware and software implementation techniques for TM. Multiple research TM compilers have been released, allowing programmers to start experimenting with transactional programming. These implementations, however, differ in their choice of transactional constructs, language restrictions and transactional semantics, thus making it difficult for programmers to write portable transactional code. To make TM a practical reality, it is necessary to come up with a standard set of transactional language constructs for major programming languages and fully define how these constructs interact with the rest of a language.
This talk will describe an ongoing effort by researchers from Intel, Oracle, IBM and HP to define transactional language constructs for C++. It will give a comprehensive overview of the released Draft Specification of Transactional Language Constructs for C+, explain the rationale behind our decisions, and give a preview of the changes upcoming in the next version of the specification.
OSF Brown Bag: Evaluating Cloud Computing for HPC Applications
Tuesday, Nov. 23, 12:00–1:00 pm, OSF 238
Lavanya Ramakrishnan, Advanced Computing for Science Department, LBNL CRD
Link of the Week: How Cats Drink
It has taken four highly qualified engineers and a bunch of integral equations to figure it out, but we now know how cats drink. The answer is: very elegantly, and not at all the way you might suppose. The cat’s lapping method depends on its instinctive ability to calculate the balance between opposing gravitational and inertial forces. And the lapping frequency equals the weight of the cat species, raised to the power of minus one-sixth and multiplied by 4.6. This is all detailed in a paper in Science Express, or you can see photos and video in the New York Times.
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.