SC09 Bandwidth Challengers Supported by ESnet
 |
 |
November 6, 2009
 |
Contacts: Linda Vu, lvu@lbl.gov, 510-495-2402
As scientists in a wide variety of disciplines increasingly rely on supercomputers and collaboration with colleagues around the world to advance their research, managing and sharing the mountain of data generated by their investigations in a timely manner is becoming increasingly difficult. Inspired by this situation, the SC09 Bandwidth Challenge provides a competition in which participants are pushed to find creative techniques and technologies for transmitting as much data as possible in a fixed period of time.
Two teams in this year's competition will be transporting terabytes of data across the Department of Energy's Energy Sciences Network (ESnet) over a period of several hours. To ensure that the data arrives within the challenge timeframe, the teams used ESnet's On-Demand Secure Circuit and Advance Reservation System (OSCARS) to reserve bandwidth on its Science Data Network (SDN), which is optimized to handle massive datasets.
"We needed a good network that would reliably move these datasets at high speeds to the convention center. ESnet was the perfect fit. The combination of SDN and OSCARS guarantees that we will have a dedicated circuit on the network for the duration of the challenge and don't have to compete with anyone else for bandwidth," says Rajkumar Kettimuthu, a staff scientist at Argonne National Laboratory and technical lead of one of the bandwidth challenge entries.
The Challengers
Kettimuthu's team will attempt to stream 10 terabytes of climate research data from three DOE computing facilities — the Argonne Leadership Computing Facility (ALCF), the National Energy Research Scientific Computing Center (NERSC) at the Lawrence Berkeley National Laboratory, and Lawrence Livermore National Laboratory — to the SC09 exhibition at the Oregon Convention Center in Portland — in a period of two-hours. Once the data arrives at the University of Utah's SC09 booth, it will be stored on disks, processed in real time with Climate Data Analysis Tools and the Visualization Streams for Ultimate Scalability visualization tool, and then publicly displayed along with graphs depicting the demo's transfer rates.
In the next few years, climate researchers will be moving terabytes of data to collaborators across the globe for the Intergovernmental Panel on Climate Change Fifth Assessment Report (AR5), which will be published in 2013. This demonstration will highlight the tools and services that will help them transport their data quickly and reliably.
Another team led by Harvey Newman of the California Institute of Technology (Caltech) will demonstrate storage-to-storage physics dataset transfers of up to 100 Gbps (gigabits per second) sustained in one direction, and well above 100 Gbps in total moving bidirectionally, using a total of 15 10-Gbps drops at the Caltech booth. The demonstration will make use of Hadoop and dCache storage systems in Portland and at partner institutes in Michigan, Florida, San Diego, Brazil and Korea, as well as the SLAC, Fermi, and Brookhaven national laboratories, and the European Organization for Nuclear Research (CERN).
High-energy physicists will break new ground when CERN's Large Hadron Collider (LHC) is scheduled to begin producing collisions later this year. The experiments are expected to produce massive amounts of data, of the order of 10 petabytes per year in 2010 and far more in later years. In order to fully exploit the potential wealth for scientific discoveries, a global-scale grid system has been developed that aims to harness the combined computational power and storage capacity of 11 major Tier 1 centers and 120 Tier 2 centers sited at laboratories and universities throughout the world. This demonstration will preview the efficient use of long-range high capacity networks that are at the heart of this system.
Beyond the Challenge
In a demonstration unrelated to the Bandwidth Challenge, a team led by Michael Norman and Rick Wagner of the San Diego Supercomputer Center (SDSC) and a group of ALCF researchers, will be using an application from the StarGate project to stream real-time visualization over ESnet's SDN from ALCF's Eureka analytics cluster to the SDSC booth on the SC09 exhibit floor. The data being visualized at Argonne is from a cosmology simulation conducted on 16,384 processor cores of the National Institute for Computational Sciences' "Kraken" Cray XT5 system. The experiment measured the imprint of matter fluctuations (baryon acoustic oscillations) caused by sound waves being frozen in place during the recombination of the early universe. In addition to exploiting the development of dynamic 10 Gbps circuits on ESnet's SDN, the demonstration will also explore use of OptIPortals as petascale supercomputer end stations for large data transfers.
For more information about Kettimuthu's challenge entry, please visit: http://www.lbl.gov/cs/Archive/news110909a.html
For more information about Newman's challenge entry, please visit: http://supercomputing.caltech.edu/index.html
For more information about OSCARS, please visit:
http://www.lbl.gov/cs/Archive/news073109.html
About ESnet and the Berkeley Lab
ESnet is primarily funded by DOE's Office of Science, one of the nation's largest supporters of scientific research. Managed and operated by the ESnet staff at Berkeley Lab, ESnet provides high-bandwidth network connections to more than 40 sites conducting DOE-funded research, including some 20 large-scale experimental facilities and large supercomputing centers used by thousands of DOE scientists generating massive amounts of data. One goal of the project is to provide a 100 Gbps link between DOE's largest unclassified supercomputing centers in California, Illinois and Tennessee.
Berkeley Lab is a U.S. Department of Energy national laboratory located in Berkeley, California. It conducts unclassified scientific research and is managed by the University of California for the DOE Office of Science. Visit our website at http://www.lbl.gov or learn more about Berkeley Lab Computing Sciences at www.lbl.gov/cs.
About SC09
SC09, sponsored by ACM (Association for Computing Machinery) and the IEEE Computer Society, offers a complete technical education program and exhibition to showcase the many ways high performance computing, networking, storage and analysis lead to advances in scientific discovery, research, education and commerce. This premier international conference includes a globally attended technical program, workshops, tutorials, an exhibit area, demonstrations and hands-on learning. The SC conference series is among Tradeshow Week magazine's Top 200 events. For more information on SC09, please visit http://sc09.supercomputing.org/.
