ACADEMIA

Scientific research and grid computing at universities across the country took a big step forward recently with an award of $10 million from the National Science Foundation to the Data Intensive Science University Network. DISUN will allow over 200 physicists at U.S. universities to study the fundamental properties of particles and forces by providing access to data from the Compact Muon Solenoid experiment at the Large Hadron Collider at CERN in Geneva, Switzerland. DISUN is a collaboration of five U.S. universities that will develop, deploy and operate a distributed computing infrastructure that will allow university physicists to analyze data collected on another continent, and researchers in other sciences to strengthen participation in national and international research activities. "Through DISUN, the physics community will be able to take full advantage of the unprecedented opportunities for discoveries at the LHC," said Harvey Newman from the California Institute of Technology, US CMS Collaboration Board Chair. "DISUN will develop a next-generation grid infrastructure that leverages advances in middleware, analysis environments and networking from universities, laboratories and grid projects worldwide." Funding for DISUN is the result of breakthrough collaboration between NSF's Directorate for Mathematical and Physical Sciences and its Directorate for Computer and Information Science and Engineering. Each directorate will provide $5 million over five years to the project, funds to be administered by the University of California, Los Angeles. "This is a critical investment for all of the physical sciences, even if some fields don't realize it yet," said Michael Turner, Assistant Director in the MPS Directorate at NSF. "For particle physics, the future of the field hangs on the LHC, where we expect fantastic discoveries from the CMS and ATLAS experiments. The data will be collected in Europe, but the grid will make it possible for U.S. graduate students, postdoctoral researchers and faculty to be close to the data so that they can make many of the exciting discoveries." Deborah Crawford, Deputy Assistant Director in the CISE Directorate at NSF, stressed the applicability of DISUN to broader communities. "One of the most exciting aspects of DISUN is that the technology and tools developed have the potential to impact many disciplines and to encourage further collaborations," said Crawford. "With support for and coordination of projects like DISUN, NSF moves toward creation of an interoperable cyberinfrastructure that will accelerate discovery, learning and innovation across all science and engineering fields." The DISUN infrastructure will build on previous grid computing and networking research also enabled by NSF funding. "DISUN is a wonderful example of how research funded through NSF's Information Technology Research for National Priorities and NSF Middleware Initiative programs is now being translated into a national cyberinfrastructure," said Sridhara Dasu of the University of Wisconsin-Madison. The four universities that will contribute computing, storage, network, middleware and personnel resources to DISUN--Caltech, the University of California, San Diego, the University of Florida and the University of Wisconsin-Madison--are part of the tiered U.S. computing infrastructure for the CMS experiment, one of four experiments currently being built at CERN in Geneva, Switzerland. When the LHC begins operating in 2007 as the world's highest energy accelerator, data will be collected at CERN and sent over high-speed networks to large Tier 1 computer facilities worldwide, including Fermi National Accelerator Laboratory in Illinois. Fermilab will provide CMS data to seven Tier 2 computing facilities at universities across the U.S., using the grid infrastructure that DISUN will develop and implement. "The petabytes of data every year that LHC experiments will produce challenge even the largest computing systems," said Fermilab's Lothar Bauerdick, US CMS head of software and computing. "We look to grid computing to provide researchers worldwide access to LHC data, and to DISUN to develop the grid tools necessary to connect U.S. universities to CMS data." Access to grid computing resources will be relevant to researchers and students in a wide range of scientific and education communities. DISUN will make such advances possible through collaborative research and technology development. "DISUN's distributed grid computing and optical networking resources will provide a powerful testbed for computer scientists, developers of grid middleware, network specialists, physicists, scientists from other fields and students in universities and laboratories across the country," said Paul Avery from the University of Florida. The grid infrastructure developed by DISUN will reach the broader scientific community through the project's involvement in university-wide, nationwide and worldwide projects such as campus grids, the Open Science Grid and the worldwide LHC computing infrastructure. "The diverse membership of the Open Science Grid will have access to the university-based cyberinfrastructure that will allow individual scientists to connect to grid resources," said DISUN technical lead Frank Würthwein from the University of California, San Diego. "We hope that within two years, scientists and students at universities across the country will be using grid resources to attack new problems and discover innovative solutions."