Volume 7, Issue 1 -
Spring/Summer 1999

Volume 6, Issue 3
Fall 1998

Volume 6, Issue 2
Spring/Summer 1998

Volume 6, Issue 1
Winter 1998

Volume 5, Issue 4
Fall 1997

Volume 5, Issue 3
Summer 1997

Volume 5, Issue 2
Spring 1997

Volume 5, Issue 1
Winter 1997

Volume 4, Issue 4
Fall 1996

Volume 4, Issue 3
Summer 1996

Volume 4, Issue 2
Spring 1996

Volume 4, Issue 1
Winter 1996

Volume 3, Issue 4
Fall 1995

Volume 3, Issue 3
Summer 1995

Volume 3, Issue 2
Spring 1995

Volume 3, Issue 1
January 1995

Volume 2, Issue 4
October 1994

Volume 2, Issue 3
July 1994

Volume 2, Issue 2
April 1994

Volume 2, Issue 1
January 1994

Volume 1, Issue 4
October 1993

Volume 1, Issue 3
July 1993

Volume 1, Issue 2
April 1993

Volume 1, Issue 1
January 1993

CRPC: THE NEXT FIVE YEARS AND BEYOND

Ken Kennedy, Director, CRPC


The National Science Foundation's announcement that the renewal of the Center for Research on Parallel Computation (CRPC) has been extended to the year 2000 prompts me to consider the role that the CRPC will play over the next five years and beyond. Since its inception in 1989, the CRPC has seen many changes, and will continue to create and meet new challenges into the next century.

The CRPC began primarily as a coordinated research program aimed at making parallel computing usable by scientists and engineers. Through the years, the Center broadened its mission to include service to the community in the form of computer network infrastructure-building and educational programs. These changes were incorporated as we realized the CRPC could not achieve its mission without educating potential users about developing CRPC technologies.

In addition, the National Science Foundation (NSF), encouraged by Congress, came to view Science and Technology Centers (STCs) as good vehicles to help fulfill its mission of involving scientists and engineers more directly in K-12 education. The NSF has also provided budgetary supplements that have encouraged STCs to develop K-12 education programs.

The CRPC has responded with educational programs that reach a full spectrum of students, including K-12 and university students, and practicing computer scientists who need to learn about parallel computation. The CRPC has also made efforts to reach the broadest ethnic and gender population at every level, as well.

Under Professor Richard Tapia's leadership, the CRPC has developed a series of "awareness workshops" for teachers, counselors, and principals of K-12 schools, particularly those with a large population of minority students. These workshops convey information about activities and careers in computational science and engineering. They also equip teachers to encourage their minority students to pursue science and engineering careers. These CRPC workshops have been extremely successful both at Rice (directed by Professor Tapia) and at Caltech (directed by Professor Herb Keller).

Building on its success with minority education programs, the CRPC is now developing programs aimed at increasing the participation of women in computational science and engineering. At the K-12 level, the CRPC developed a four-week course for technology teachers that covers both the use of the World Wide Web and material on gender issues. This course exposes technology teachers to the power of the computer as an information tool, while conveying an understanding of why middle school and high school girls are often left out of computing activities. The first session in this program, called GirlTECH '95, was held in Houston this summer under the direction of Cynthia Lanius, Associate Director of the Rice University School Mathematics Project (RUSMP) and CRPC Project Manager, GirlTECH '95.

Further, the CRPC has initiated two new programs aimed at technology transfer and education for the end user. The first is the National HPCC Software Exchange (NHSE), which was initiated at CRPC with funding from five federal agencies led by NASA. The NHSE is building a World Wide Web-based repository and dissemination system for HPCC software, particularly software developed under the federally-funded HPCC research program. The prototype NHSE has been available since November 1994. The NHSE home page can be accessed at http://www.netlib.org/ .

This summer, the NHSE will begin its software review program, in which selected software will be certified to be "partially reviewed" or "reviewed." Partially reviewed software will be inspected by NHSE librarians and certified to achieve a certain level of completeness. Reviewed software will be examined and executed by two or more researchers to determine that it is functional and usable. In the future, NHSE may add categories that represent even more rigorous examinations.

The NHSE will also include a detailed "road map" designed by HPCC experts to help the user find the right software quickly. Currently, a prototype of the NHSE road map, also based on World Wide Web technology, can be reached from the NHSE home page.

The second CRPC educational program for end users is the "National Retooling Project." It teaches existing supercomputer users about parallel computation. Because the CRPC does not have sufficient resources to educate large numbers of current supercomputer users, it has focused on producing educational materials and courses that supercomputer staff trainers can use to teach their users. This effort, funded by an NSF Metacenter Regional Affiliates program, collaborates with the North Carolina Supercomputer Center and the four Metacenter sites.

During the first year, course materials were developed on MPI, Fortran M, templates for sparse iterative methods, and business/science applications of parallel computing. This program provides an ideal way for the CRPC to disseminate information about parallel computing technologies to the science and engineering communities.

In addition to new educational programs, the CRPC has been evolving its research program to keep pace with the changing world of parallelism. Over the past two years, it has initiated new programs on scalable input/output, parallelization of engineering computation, templates and archetypes for parallelism, and parallel computing on clusters of multiprocessor workstations. Over the next five years, it will continue to adapt its programs in scientific computing as well as in education/outreach to capitalize on its successful efforts in both areas to date.

The CRPC's directions both lead and reflect trends in our country's overall goals for high-performance computing and communications for the next five years and beyond. At a recent Committee on Information and Communications (CIC) Forum in Washington, D.C., 140 industry, academic and government representatives-including CRPC leaders-considered new technologies, outlined new research directions, and reshaped HPCC initiatives for upcoming years. Both CRPC and the larger HPCC community recognize that we need to understand how we can use HPCC technologies in education as well as how to educate people to use the technologies.

The government has kept close watch over the NSF Centers programs, and every systematic review has validated their valuable role in science and engineering research. For this reason, it is likely that there will be a new Science and Technology Center competition in 1997-98. If so, the team that built the CRPC plans to submit a new proposal in a different but related area. I have been privileged to be associated with the group of researchers and staff members that together make up the CRPC, and I am confident that our vision will extend far into the next century.


Table of Contents