InTheLoop | 06.25.2012
June 25, 2012
NERSC Signs Supercomputing Agreement with Cray for Next-Gen System
NERSC announced that Cray will install its next-generation supercomputer computing system. Consisting of products and services, the multi-year, multi-phase project is valued at more than $40 million. It includes delivery of a future-generation Cray supercomputer code-named "Cascade" and a next-generation Cray Sonexion storage system. When completed, the new system will deliver a peak performance of more than two-petaflops per second, equivalent to more than two-quadrillion calculations per second. The full system is expected to go into production in 2013. Read the full announcement.
The Register, a global online tech publication with offices in London and San Francisco that reaches more than six million unique users worldwide, highlighted the announcement. Read the article.
Science at NERSC: Sifting Through a Trillion Electrons
Using NERSC resources, astrophysicists generated a trillion-particle magnetic reconnection dataset in 3D, where each time-step of the simulation amounted to a massive 32 terabyte file. Armed with new tools developed by researchers in Berkeley Lab's Computational Research Division (CRD), the scientists were able to query this enormous dataset for particles of interest in three seconds, and visualize it. This is the first time a dataset of this magnitude has been queried and visualized this quickly. Read the full story.
University of Utah Implementing ESnet's Science DMZ Architecture
Ars Technica, a publication devoted to technology that describes itself as catering to "alpha geeks," or technologists and IT professionals, recently published an article on how the University of Utah is deploying the Science DMZ architecture developed by ESnet to speed the flow of data.
The article begins: "Thanks to networks like the U.S. Department of Energy's ESnet, scientists crunching huge bodies of data finally have 10Gbps pipes at the ready to zap that information to their peers anywhere in the world. But oftentimes these blazing speeds are torpedoed by firewalls and other security devices. So using technology developed or funded by NERSC, ESnet, NSF and others, the University of Utah set out to architect a new security design, called 'Science DMZ,' that doesn't put a crimp on bandwidth." Read the full article.
This Week’s Computing Sciences Seminars
Dynamic Adaptation for Service Based Applications
Tuesday, July 3, 2 p.m.-3 p.m.
Bldg. 50B, Room 4205
Jean-Louis Pazat, Myriads Team IRISA/INSA, Rennes, France
Nowadays, software design often relies on the composition of services rather than on components. This composition is often statically defined by a workflow representing a business process. There is no strong coupling between services which are under the responsibility of different service providers. As this process may evolve during execution due to some errors, malfunctions, faults of services or even changes in the business process; the composition of services and the services themselves should be able to change their behavior during execution.
These changes, or "Dynamic Adaptation," are becoming a standard in software design and are often implemented with the MAPE model. Opposite to High Performance Computing where people always try to get the best performance out the computing resources, in most cases for service based applications (SBA), we do not try to optimize the Quality of Service (QoS) but to conform to some negotiated QoS. The negotiation process relies on SLA (Service Level Agreements). When service execution relies on a cloud computing platform, service adaptation can be achieved at different levels (application level, platform level of infrastructure level). The big challenge here is to avoid conflicting adaptations at different levels; which implies to deal with multilevel adaptation.
In this talk I will show some ideas developed in the S-Cube European Network of Excellence and frameworks developed in the Myriads Teams for dynamic adaptation of distributed applications and their use in the context of Service Based Applications. I will present some challenges of multi-level adaptation for the execution of SBAs on Clouds and show some ideas for dynamic resource provisioning based on SLAs. In conclusion I will also show an advanced programming model suited for dynamic adaptation.
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 7,000-plus 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 Department of Energy 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.