The team that developed I-Soft, the middleware used for the I-WAY experiment at Supercomputing 95, has been working for the past two years on the design and implementation of Globus, an integrated set of software components for next-generation high-performance internets. Globus takes major steps forward in security, scheduling, performance, and usability to allow, for the first time, the truly coordinated use of supercomputers in different states and on different continents.

Globus software will premiere at the SC'97 conference in San Jose, California, where it will be demonstrated on a new metacomputing testbed of unprecedented scale and power. The Globus Ubiquitous Supercomputing Testbed (GUSTO), created by a multi-institutional team led by Globus developers at Argonne National Laboratory and the University of Southern California's Information Sciences Institute (ISI), links over 3000 processors at 13 sites in the U.S. and Europe. GUSTO achieves a high degree of integration between its diverse sites by using Globus software for authentication, resource location, scheduling, communication, and fault detection.

Globus researchers view GUSTO as a prototype for future "computational grids," which provide computer power to researchers for problem solving, just as a power grid provides electricity to people at home or at work. Such grids will place the most advanced supercomputers, data archives, virtual-reality displays, and scientific instruments at the fingertips of the nation's scientists and engineers--regardless of where tools or people are located--and hence enable new problem solving techniques, such as distributed supercomputing, remote visualization, and tele-immersion.

"By providing pervasive access to supercomputing capabilities, computational grids will change the way we think about and use high-end computing," said Dr. Ian Foster of Argonne, who co-leads the Globus project with Dr. Carl Kesselman of ISI. "We're excited to be taking this step towards the establishment of a permanent computational grid facility."

"From previous experiments, we know that both the technical and organizational obstacles to creating such integrated grids are tremendous," said Kesselman. "However, the cooperation that we have received from all participants has been amazing. People are clearly ready for this next step towards wide-spread collaborative supercomputing."

The GUSTO grid was developed in collaboration with the National Computational Science Alliance (Alliance) and the National Partnership for Advanced Computational Infrastructure (NPACI), the recipients of the NSF Partnerships for Advanced Computational Infrastructure (PACI) program, as well as with staff at DOE and NASA laboratories and other research centers around the world. The Alliance, anchored by the National Center for Supercomputing Applications (NCSA) at the University of Illinois at Urbana-Champaign, helped develop essential software and incorporated its large SGI/Cray Origin2000 and Convex supercomputers into the grid. NPACI, anchored by the San Diego Supercomputer Center (SDSC) at the University of California, San Diego, is also supporting the effort and providing access to computers.

Other sites participating in GUSTO include the California Institute of Technology, the Paralleldatorcentrum at Kungliga Tekniska Hogskolan in Sweden, Indiana University, the National Energy Research Scientific Computing Center, Los Alamos National Laboratory, NASA Ames Research Center, the Rechenzentrum Garching der Max-Planck-Gesellschaft Garching in Germany, the Texas Center for Computational and Information Sciences at the University of Houston, the Maui High Performance Computing Center, and the Condor Project at the University of Wisconsin. GUSTO resources are connected by a variety of high-speed networks, including ESnet, vBNS, and international networks accessed via STAR TAP.

One novel resource incorporated in the GUSTO grid is a large pool of workstations managed by Condor, a system developed at the University of Wisconsin by Prof. Miron Livny and his colleagues to support high-throughput computing. "We're excited to be coupling Condor with Globus," said Livny. "We believe that future grids will need to support both high-performance and high-throughput computing, and this seems to be the way to do it."

At the SC'97 conference, 10 groups will use Globus software and GUSTO resources for a range of exciting distributed supercomputing applications, including:

• SF-Express, a distributed interactive simulation system developed by Sharon Brunett, Paul Messina and others at Caltech and JPL; this uses multiple supercomputers to perform very large synethetic forces simulations

• A system that uses remote supercomputers for the reconstruction of crystallographic data from the Advanced Photon Source at Argonne, developed by Gregor von Laszewski, Mary Westbrook, and others

• Neph, a supercomputer-enhanced instrumentation application developed by Craig Lee of The Aerospace Corporation, that uses dynamically acquired computing resources for analysis of satellite data

• NICE/CAVERN, a system developed by Jason Leigh and others at the Electronic Visualization Laboratory at the University of Illinois at Chicago, for the creation of shared virtual spaces

• Cactus, a system for computing and visualizing Einstein's gravity wave equations at a distance, developed at the Max-Planck Institute in Potsdam, Germany, at NCSA, and at Washington University St. Louis. (This application will be run across the Atlantic, using a supercomputer in Garching, Germany.)

• Nimrod, a system for creating and managing large parameter studies, developed at Monash University in Australia; this will be used for environmental and superconductivity studies

• MPMM, an integrated weather modeling system that will use remote computing resources to generate daily weather forecasts during the SC'97 conference

Among the events planned for the SC'97 conference is an attempt at a world-record SF-Express run. The goal is to harness a sizeable fraction of GUSTO resources to achieve a simulation involving 60,000 entities.

Globus research and development is supported by DARPA, DOE, and NSF, and by an equipment grant from Sun Microsystems. For more information concerning Globus, contact Ian Foster at Argonne National Laboratory at foster@mcs.anl.gov, +1.630.252.4619, or Carl Kesselman at ISI at carl@isi.edu, +1.310.822.1511, or visit http://www.globus.org/.

Contact:
Ian Foster
foster@mcs.anl.gov
ph: +1.630.252.4619

Carl Kesselman,
carl@isi.edu
+1.310.822.1511