APPLICATIONS

CHAMPAIGN, IL--A new program at the National Center for Supercomputing Applications (NCSA) will help scientists find solutions to their toughest computational problems so they are better able to achieve their research objectives. The Strategic Applications Program (SAP) also is expected to benefit the larger scientific computing community by solving problems related to scaling, I/O, cache optimization, porting codes to different platforms, and a wide range of other issues. "This program involves hands-on collaboration between research teams running applications on our systems and technical staff here at NCSA in an effort to find solutions to the problems that often arise when you are involved in computing at this level," said John Towns, director of NCSA's Scientific Computing Division. "Even more important, these projects could result in new methodologies, algorithms, or other solutions, which we will then share with the broader user community." Located at the University of Illinois at Urbana-Champaign, NCSA provides compute resources to hundreds of scientists in fields ranging from molecular biology to cosmology to atmospheric science. Each year, these researchers use millions of hours on some of the world's most powerful supercomputers. Their work involves analysis of massive amounts of data, complicated algorithms that measure a multitude of real-world phenomena, and high-resolution visualizations that show changes over time and space. Such work often pushes the limit of supercomputer systems and demands development of better algorithms, memory hierarchy systems, performance analysis tools, and tools that allow for better parallel processing and distributed computing over grids. The SAP will address all these issues and more so that research teams will have better performance on NCSA systems and be able to conduct more research. NCSA senior management and the SAP coordinator will chose projects for the program, and each project will be assigned an NCSA Strategic Application Project Lead. Most SAP projects will last about six months and will be chosen based on the following criteria: * a high-quality research project, as evidenced by publications, grants, and honors * a clear need for powerful resources for computation, visualization, and/or processing large data sets * a well-defined and attainable set of goals for the intended time frame * a clear possibility that the results of the SAP will benefit a larger community of users * a willingness by the research Principal Investigator and team members to collaborate with NCSA staff. For more information, see http://www.ncsa.uiuc.edu/UserInfo/Alliance/SAP/ .