Complex computing problems as different as modeling Earth's climate system, predicting effects of regulatory change in the dairy industry or serving a semester's worth of lecture videos to student dormitories will operate on a scalable distributed network of powerful desktop computers, thanks in part to a $6 million grant from Intel Corp. to Cornell .

The grant from the Santa Clara, Calif., computing equipment manufacturer is one of 12 to American universities in Intel's three-year, $85 million "Technology for Education 2000" program to support research and curriculum development in computationally demanding fields.

"This $6 million grant from Intel will help create a new learning environment at Cornell," said University Provost Don M. Randel. "This environment will make it easier for people to work together as effective teams, it will engage students in problems with 'real world complexity' and it will support multiple styles of learning."

Building a scalable distributed system is the best way to serve thousands of clients with complex computing needs, according to Daniel P. Huttenlocher, associate professor of computer science and one of the planners of the proposed network at Cornell.

"We intend to use high-speed communications networks and substantial desktop and server computing power to deliver information in ways that are more visual, more customizable and more interactive than has ever been attempted in an academic setting," Huttenlocher said. "We will create 'anything-anytime-anywhere' access to information by decoupling location and function."

"Students and faculty will be able to collaborate on projects across campus, access lectures from dormitories and search library materials from any desktop computer," said Dean B. Krafft, director of computing facilities for the Department of Computer Science, noting that Cornell was among the first universities to wire every dormitory room with high-speed network connections and now is building a 150-Mbit switched campus "backbone."

The Intel grant will provide selected Pentium II-based processors and workstations, software and training as well as upgrades to more powerful Intel processors as they reach the marketplace. The scalable distributed computing system will support a variety of curriculum-development and research projects (see attached list of examples) for students and faculty members in five university units: the College of Engineering, the College of Agriculture and Life Sciences, the S.C. Johnson Graduate School of Management, the College of Arts and Sciences and the College of Architecture, Art and Planning.

Commenting on the grants, Intel's Corporate Contributions Manager Tim Saponas said, "The long-term success of our industry and other industries depends on the quality of U.S. research universities, their computing infrastructure and the skills of their students. We are pleased to be able to support some of the best universities in the country with state-of-the-art tools to carry out computationally intensive research projects." Also receiving Intel grants are the universities of California at Berkeley, Michigan at Ann Arbor, Texas at Austin, Wisconsin at Madison, Southern California and Washington, California Institute of Technology, Carnegie Mellon University, Georgia Institute of Technology, Massachusetts Institute of Technology and Purdue University.

"If we are successful in achieving our most ambitious goal, of fundamentally transforming the ways in which information is created and disseminated in the university," Provost Randel said, "then Cornell will serve as a national model for universities in the information age."

Examples of computational, curriculum-development and research projects planned with Intel grant to Cornell

• Create a scalable distributed computation resource (SDCR): Servers and desktop machines based on Intel Architecture (IA) will become integrated parts of a massive computing environment. Using the Advanced Resource Management System (ARMS), computational jobs will run on appropriately scaled resources, ranging from a small number of nodes in a departmental cluster to additional machines in divisions, colleges, all the way to the total computational resource composed of hundreds of IA processors.
• Molecular dynamics simulations in chemical engineering: Researchers plan to use the SDCR system for massive computational tasks such as predicting macroscopic properties of fluids and materials from molecular-level information.
• Environmental Computing Laboratory: Climate system modeling will help researchers understand the relationships between land and sea surface conditions and the features of atmospheric dynamics and precipitation. Digital map servers in the environmental laboratory will make dynamic, customized information available to land owners, engineers, resource managers and farmers, as well as to students and researchers.
• Interactive numerical help center for physics students: Simulation modules for interactive learning -- in such topics as electron transport, lattice vibrations, harmonic motion and two-dimensional wave interference -- will be accessible to more than 2,000 freshman and sophomore students in engineering physics and pre-medical physics courses on an anything-anytime-anywhere basis.
• Computational biology parallel processing cluster: A 56-node parallel processing Pentium cluster will include high-performance graphics workstations for macromolecular modeling and biophysical imaging. The cluster also will participate in the SDCR to provide the most efficient distributed use of computing cycles across campus.
• Predicting effects of technological and regulatory change: Researchers in the Cornell Program on Dairy Markets and Policy can analyze impacts of changes in technology, demographics, economics and the regulatory environments to advise policy-makers, industry leaders and individual producers.
• Next generation distributed digital library: Beginning with the more than 27,000 works housed in Cornell's Johnson Museum of Art, the digital library plans to capture the full range of intellectual content housed and generated at the university, from technical papers and reports to bird songs, crop databases and new works of digital art--
• Solving QCD: Physicists propose to use a small number of Intel's most powerful desktop workstations for a task has been the province of the largest supercomputers -- solving quantum chromodynamics (QCD), the fundamental theory of proton substructure.
• Investment research: Intel equipment in the Parker Center for Investment Research Initiatives of the Johnson Graduate School of Management will integrate real-time data on stock prices and quotes from global markets to perform financial analyses for a student-managed investment fund, operating from a state-of-the-art trading room in the renovated Sage Hall.