Advances in science and engineering are driven increasingly by collaborations that focus on sharing data, computing, code,
and access to experimental facilities. Network-driven computers, storage, data collections and scientific instruments
are now central to the day-to-day practice of many research disciplines.
For example, the National Science Foundation’s GriPhyN project uses an international network of computational systems and
data collections to address next-generation particle physics experiments at the Large Hadron Collider (LHC), while the NSF-
funded George E. Brown Network for Earthquake Engineering Simulation (NEES) is revolutionizing seismology via network
enabled access to experimental facilities, data, and simulations. GriPhyN and NEES represent not today’s standard practice,
but five- to ten-year strides for these disciplines. GriPhyN is preparing for a torrent of data from LHC experiments to
begin in 2006 with a 15-year duration, while NEES is expected to be in place until 2014.
Though these communities are ready now to develop
new modes of research, scientists and engineers are
frustrated by the scarcity of network-enabled services
to suit their applications. The NSF
Middleware Initiative has begun as an effort to lead
the way toward next-generation infrastructure for
large-scale, flexible resource sharing on national and
international scales. The Grid Research Integration Development and
Support (GRIDS) Center has been created through NMI to define, develop, deploy, and support an integrated national
middleware infrastructure in support of 21st Century science and engineering applications. GRIDS is a partnership
of the University of Southern California’s Information Sciences Institute (ISI), the National Center for Supercomputing
Applications (NCSA) at the University of Illinois at Urbana-Champaign, the University of Chicago (U of C), the
San Diego Supercomputer Center (SDSC) at the University of
California-San Diego and the University of Wisconsin-Madison.