BIG DATA

The National Partnership for Advanced Computational Infrastructure (NPACI) is emphasizing "big" in a research exhibit and related demonstrations at SC2002, this year’s edition of the world’s largest annual high-performance networking and computing conference. Visitors to the NPACI exhibit booth will have not only a chance to explore supercomputing technology and the breakthrough science it makes possible, but they also are invited to participate in the community that makes up NPACI, which includes 41 partner institutions in 17 states, and Australia, Italy, Spain, and Sweden. With the theme "From Terabytes to Insights," SC2002 will be held November 16-22 at the Baltimore Convention Center. "The SC2002 NPACI booth showcases big science, big technology and big infrastructure," said Fran Berman, director of NPACI and the San Diego Supercomputer Center, NPACI's leading edge site at UC San Diego. "This is a great year for the NPACI partnership. Our projects illuminate how infrastructure can be used to advance science and enable new discoveries. The NPACI booth has something for everyone, whether you are interested in enabling technologies such as grid computing, data and knowledge management, networking, and high-end computing, or computational science advances in astrophysics, bioinformatics, chemistry, geoscience, and medicine. This is a unique opportunity to sample NPACI at its best and interact with some of the most outstanding scientists and technologists in the world." Along with highlighting the accomplishments of NPACI's projects, the partnership’s booth will showcase its role in the TeraGrid, which will be fully deployed in 2003 as the world's most powerful distributed computing infrastructure for open scientific research. Also prominently highlighted in the NPACI exhibit booth will be the Biomedical Informatics Research Network, an initiative to share high-resolution brain images and other biomedical data across the country over a high-speed communications network, and the Pacific Rim Applications and Grid Middleware Assembly (PRAGMA) multinational collaboration. Researchers will demonstrate how NPACI-developed technology has led to major scientific successes, such as visualizing the evolution of a galaxy, simulating the flow of blood through a heart, and the creation of a National Virtual Observatory. The schedule of NPACI demonstrations is at http://www.npaci.edu/sc2002/ and include: ** The Ultimate Protein Database: The Encyclopedia of Life (EOL) is an ambitious project to catalog and annotate the "proteome" of every investigated species of life and make this data available through an advanced Web interface and Web services. The EOL will give biomedical researchers a one-stop location for protein sequence annotations, comparative data, and structural predictions. The software runs on computing platforms that range from Linux clusters to the IBM Blue Horizon supercomputer at SDSC. The TeraGrid system will soon give the EOL the computing resources needed to analyze a vast amount of genomic data. ** Supercomputer-Enabled Surgery: SC2002 Plenary Speaker Ron Kikinis of Brigham and Women's Hospital and Harvard Medical School will explain how image-guided therapy (IGT) allows doctors to see through tissue. ** National Virtual Observatory: Several presentations will showcase the capabilities of the National Virtual Observatory, a project to manage and mine a truly astronomical amount of data. The Montage project at Caltech's Infrared Processing and Analysis Center is developing an astronomical image mosaic service to deliver science-grade custom mosaics. One demonstration retrieves data and creates mosaics from the near-infrared Two-Micron All Sky Survey (2MASS), a four-terabyte repository stored on SDSC's storage area network TeraGrid disk array, using Montage software running on SDSC's Blue Horizon supercomputer. Other demonstrations by SDSC and Caltech researchers will create images using the Itanium 2 TeraGrid cluster. ** Cosmic Web: Michael Norman, a senior fellow at SDSC and a physics professor at UCSD, will demonstrate a data portal for cosmological simulations. The portal manages simulations of the distribution of matter in the early universe. The simulations on Blue Horizon are among the largest ever computed—a single run can generate more than one terabyte of data. The presentation features real-time, interactive volume visualizations of the results. ** Windows to the Grid: The NPACI TeraGrid portal, developed by NPACI's Grid Portals Architecture group, will let researchers create a personalized gateway to the Grid and access advanced storage resource management features. ** Oil Reservoir Simulation: Joel Saltz of Ohio State University and Alan Sussman of the University of Maryland will explore and visualize ensembles of oil reservoir simulations. Their method provides an efficient and cost-effective means for accurate characterization of underground oil reservoirs. ** Interactive Telescience across the Globe: The National Center for Microscopy and Imaging Research (NCMIR) at UCSD, in collaboration with Osaka University in Japan and Taiwan's National Center for High-Performance Computing (NCHC), is demonstrating remote control of an intermediate-high-voltage electron microscope in San Diego and an ultra-high-voltage electron microscope in Osaka. The demo features real-time control and HDTV and digital video over high-performance IPv6 networks. Researchers will use the Telescience system to explore a specimen in a microscope, create data-intensive visualizations, and transfer data among resources located at: NCMIR, SDSC, NASA Ames Research Center's Information Power Grid, NCHC, and Osaka University. ** International Collaboration Focused on Grid Application: Research from several PRAGMA members will describe this new international collaboration to advance the use of Grid technologies among leading institutions around the Pacific Rim. Summaries, locations, and times of all of the presentations and demos are featured at http://www.npaci.edu/sc2002/cgi-bin/sc2002_schedule.cgi. The SC2002 research exhibits will open at 7 p.m. on November 18, and will be open daily through November 21. More about the conference, including a daily schedule organizer, is available at http://www.sc2002.org/.