Image Courtesy San Diego Supercomputer Center

"The attendees all got an intensive introduction to the resources, techniques and tools for dealing with the extreme data and I/O needs of today's computational science applications," said David Hart, SDSC allocations manager and organizer of this year's summer institute. "By all accounts, everyone left excited and inspired, with new ideas for applying the latest capabilities in their own research."

The institute, which was funded by the National Science Foundation, covered a number of topics including: how to use the SDSC computing environment; parallel I/O; HDF5 and NetCDF file formats; moving data across the TeraGrid; building and using data collections; an introduction to database design; using the SDSC Storage Resource Broker; data mining and workflows; data visualization and case studies of extreme I/O applications.

"The great feature of this institute was that attendees actively participated in the different topics presented," said Andrea Silvestri, a graduate student in particle astrophysics at the University of California, Irvine, who presented research completed using TeraGrid facilities at the institute. "We received accounts on many SDSC machines, which gave us the opportunity to practice the examples and test the methods explained in the lectures."

"The institute really emphasized the integration of CPU cycles, mass storage and high performance networks into one efficiently functioning mechanism," added computer science graduate student Jakub Kurzak from the University of Houston. "The day is yet to come when a scientific simulation can be run with a push of a button, but the summer institute showed the light at the end of the tunnel."

View the detailed course program and presentations at the SDSC Computing Institute Web site, or visit the SDSC Web site.