Our Mission Supercomputing Resources Current Resources Laboratories of the Supercomputing Institute Collaborative Projects Other Collaborators Other Programs Department Index Principal Investigator Index

Our Mission The Supercomputing Institute for Digital Simulation and Advanced Computation is an interdisciplinary research program spanning all colleges of the University of Minnesota. The Supercomputing Institute provides supercomputing resources and user support to faculty and students and is a linchpin program in the University's broad-based digital technology effort. The mission of the Supercomputing Institute is supercomputing research. This includes all aspects of high-performance computing and scientific modeling and simulation as well as graphics, visualization, and high-performance network communications. Supercomputing research is defined broadly to include a variety of research activities from many disciplines. This research involves the use of high-performance computing environments to address problems in science and engineering that could not otherwise be attempted. Such efforts often result in domain-specific algorithms and codes that exploit the available computing environments as well as visualization techniques to enhance insight, make displays more informative, and add multimedia value to communications and work environments. In many cases, these research activities may involve research aimed at the design or evaluation of high-performance computing hardware, operating systems, networking, and general-purpose algorithms and software. The Supercomputing Institute's resources are available to researchers at the University of Minnesota and other post-secondary educational institutions in the State of Minnesota. In addition, the Supercomputing Institute organizes and hosts symposia, workshops, and seminars and coordinates other educational and collaborative activities to promote supercomputing research, increase university-industry collaboration, and promote technology transfer. Return to Top

Supercomputing Resources In 1981, the University of Minnesota was the first American University to acquire a supercomputer (a Cray-1B). The Supercomputing Institute was created in 1984 to provide leading edge high-performance computing resources to the University of Minnesota's research community. These resources have included a Cray-2, an ETA 10, a Cray X-MP, an IBM 3090, a Cray M90, a Cray T3D, a twelve-processor Cray C90, and a Cray T3E-900. These resources were provided to the Supercomputing Institute by Network Computing Services (formerly Minnesota Supercomputer Center, Inc., or MSC Inc.). Beginning on July 1, 1998, the Institute began providing services on its own supercomputers. Building upon the strong tradition of providing University of Minnesota researchers with leading edge high-performance computing technologies and diversified programs that complement these technologies, the Supercomputing Institute acquired in May 1998 a 256-processor IBM SP with 192 GB of memory. This IBM SP supercomputer is a 64-node machine with four 332 MHz 604e processors and 3 GB of memory on each node. The Supercomputing Institute also offers its researchers access to a 128-processor SGI Origin 2000 R10000 supercomputer which has a clock speed of 195 MHz and a total of 64 GB of memory. In June, the Institute upgraded 48 processors of this machine to R12000 technology. The upgraded processors have a total of 36 GB of memory. Fourteen more R12000 processors with 14 GB more memory are scheduled for delivery for a total of 62 R12000 processors with 50 GB of memory. The R12000 processor, which has a clock speed of 300 MHz, is estimated to be 1.7 times faster than the R10000 processor. With the arrival of 80 WinterHawk nodes in May 1999, the Institute acquired the lastest SP technology from IBM. Each of the WinterHawk nodes contains two 200 MHz Power3 processors sharing 1 GB of memory for a total of 160 WinterHawk processors. In October 1999, the IBM SP will be expanded to include 17 NightHawk nodes. Each node will contain 4 Power3 processors sharing 16 GB of memory for a total of 68 NightHawk processors. When fully configured, the IBM SP will contain 228 processors and 352 GB of memory. We anticipate that the IBM SP will be 2.5 times faster than the current IBM SP that contains 64 Silvernodes.The IBM supercomputer resources are available in coordination with the IBM Shared University Research (SUR) parthership. The total amount of disk available to users of the IBM SP systems is currently 1.5 Terabytes. Return to Top

Current Resources The supercomputing resources are located at the Supercomputing Institute's facilities in the Minnesota Technology Corridor at the edge of the West Bank of the Minneapolis Campus of the University. In addition to IBM SP and SGI Origin 2000 supercomputers housed at the Supercomputing Institute, the Supercomputing Institute offers IBM SP resources through its IBM East Bank Laboratory. This laboratory, which is co-managed by the Supercomputing Institute and the Computer Science and Engineering Department, includes 72 processors with 36.5 GB of memory and 500 GB of disk. These IBM and SGI supercomputers offer the Supercomputing Institute's researchers access to state-of-the-art high-performance computing technology. In addition, the Supercomputing Institute is continuing its commitment to a diversified array of computing laboratories, collaborations, and programs. These include the Basic Sciences Computing Laboratory, the Scientific Development and Visualization Laboratory, the Medicinal Chemistry-Supercomputing Institute Visualization/Workstation Laboratory, and interdisciplinary Ph.D. programs in Scientific Computing and Computational Neuroscience. Return to Top

Laboratories of the Supercomputing Institute IBM Shared University Reserach East Bank Laboratory The IBM SUR East Bank Laboratory is a cooperative research effort in scientific and parallel computing between the University of Minnesota and IBM. The project is managed under the auspices of the IBM Shared University Research (SUR) program by the Computer Science Department, the Office of Information Technology, and the Supercomputing Institute. This program makes a cluster of IBM RS/6000 workstations and an SP POWERparallel system available to University of Minnesota researchers for research on system software, communications, parallel algorithms, and computational science applications. As part of the program, the University has an IBM RS/6000 Cluster, which consists of four subsystems: an 8-node RS6000 workstation subcluster, a 16-node SP supercomputer, a 4-node, 32-processor SMP supercomputer, and a 4-node, 16-processor SMP Silvernode supercomputer. The total number of processors available is 72 and the total amount of memory is 36.5 GB. In addition, there are several machines that support the RS6000 Cluster. These include an RS/6000 G30 that acts as a file server, an RS/6000 model 380 and an RS/6000 43P for staff use, and the SP control workstation. These computers are housed in the Computer Science building. The SUR Cluster is supported by IBM, the University of Minnesota Supercomputing Institute, the National Science Foundation through the Directorate for Computer and Information Science and Engineering, the Office of Information Technology and the Department of Computer Science at the University of Minnesota, and various University of Minnesota researchers through individual research grants. Basic Sciences Computing Laboratory In 1996ñ97, the Supercomputing Institute established the Supercomputing Institute Basic Sciences Computing Laboratory in the Nils Hasselmo Hall located on the East Bank campus of the University of Minnesota. The laboratory provides high-performance workstations and visualization equipment to enhance the research capabilities of the University community. The facility occupies approximately 1700 square feet that includes a workstation room, video/graphics room, machine room, and two offices. The laboratory houses state-of-the-art computing platforms and graphics workstations including five Silicon Graphics Indigo2 Solid Impact R10000 workstations, a Silicon Graphics Challenge XL with four 150 megahertz R4400 processors, two SGI Octane SSI workstations, one SGI Octane MXE workstation, two SGI Octane SSEs, one SGI Onyx2, one IBM Intellistation, and a 32-processor Origin 2000 with 8 GB of memory. The Supercomputing Institute provides technical support for these high-performance computing resources, and the laboratory is available to all University of Minnesota researchers. Medicinal Chemistry-Supercomputing Institute Visualization/Workstation Laboratory The Medicinal Chemistry-Supercomputing Institute Workstation/Visualization Laboratory is cosponsored by the Department of Medicinal Chemistry and the Supercomputing Institute. This laboratory is located in Weaver-Densford Hall and contains workstations that are used primarily for scientific visualization. This facility is available to all researchers who apply for and receive competitively reviewed supercomputing resource allocations. Access is also available by separate proposal for those researchers requiring workstation access only. Scientific Development and Visualization Laboratory The Supercomputing Institute's Scientific Development and Visualization Laboratory, which is located in the Supercomputing Institute's facilities in the Supercomputer Center Building, provides front-end equipment including Silicon Graphics workstations, Macintosh workstations, an Intellistation from IBM running Windows NT, a color scanner, a CD writer, a Silicon Graphics O2 workstation for the creation and manipulation of videos, and a Super-VHS VCR. The Institute provides user support services for supercomputer-related research using these general-purpose systems. The computer resources and user support services of the Scientific Development and Visualization Laboratory are available to all researchers who apply for and receive competitively reviewed supercomputer resource allocations. Access is also available by separate proposal for those researchers requiring workstation access only. National Center for Supercomputing Applications Resources The Supercomputing Institute is also coordinating the allocation of National Center for Supercomputing Applications (NCSA) SGI Origin 2000 resources to University of Minnesota faculty researchers. These resources are being made available through the Committee on Institutional Cooperation (an academic consortium of Big Ten Universities and the University of Chicago). The goal of all of these efforts is to foster state-of-the-art research, interdisciplinary collaboration, and university/industry collaboration by providing Minnesota researchers with access to advanced facilities for digital computing, visualization, and networking. Return to Top

Collaborative Projects IBM Shared University Research Workstation Award Program IBM, through its Shared University Research Program, has made available to the University of Minnesota Supercomputing Institute, a number of IBM RS/6000 and Intellistation workstations. These workstations have been awarded through a competitive grant program administered by the Supercomputing Institute. The Supercomputing Institute and the Computer Science and Engineering Department provide hardware and system support and user and parallel programming support. The Intellistation workstations are Z Pro Windows NT machines with Pentium II 450 MHz processors, 256 MB RAM, 9.1 GB hard drives, 32 way CD readers, Intergraph Intense Graphics 3400 video cards, and 10/100 ethernet connections. The workstations have a 21î monitor. The RS/6000 UNIX workstations contain two 332 MHz 604e processors, the same processor in the Silvernodes of the IBM SP and the GTX550P graphics adapter. Laboratory for Computational Science and Engineering The Supercomputing Institute partners with the Laboratory for Computational Science and Engineering (LCSE) in support of the LCSE's participation in the National Center for Supercomputing Applications (NCSA) National Computational Science Alliance, which has been funded by the National Science Foundation. Through this partnership, Supercomputing Institute researchers are able to participate in the LCSE program. The LCSE encourages the participation of Supercomputing Institute researchers with applications that demonstrate or test new technologies under active development, applications requiring very large on-line data setsóparticularly if they must be accessed at very high bandwidth, applications requiring very high-resolution visualizationsóparticularly if image animations are needed, and distributed computing applications with tight coupling of computing resources on a fast network are also encouraged. Scientific Computation Graduate Program The graduate degree program in scientific computation encompasses course work and research on the fundamental principles necessary to use intensive computation to support research in the physical, biological, and social sciences and engineering. There is a special emphasis on research issues, state-of-the-art methods, and the application of these methods to outstanding problems in science, engineering, and other fields that use numerical analysis, symbolic and logic analysis, high-performance computing tools, parallel algorithms, supercomputing and heterogeneous networks, and visualization. Scientific Computation is gradually emerging as an important field of its own in academia and industry. In the last decade, it has become clear that solving a given scientific problem often requires knowledge that straddles several disciplines. This interdisciplinary program provides a new combination of studies for solving today's scientific computational problems. It is a degree program that builds on the strength of existing programs at the University of Minnesota in formulating real problems based on the physical system or the traditional discipline, and it augments field-specific work relating to the mathematical and numerical modeling with state-of-the-art techniques for scientific computation in an integrated manner. The Scientific Computation program offers Ph.D. and M.S. degrees. Computational Neuroscience Graduate Program This program introduces students with diverse biological and quantitative backgrounds to the challenges of complex phenomena in the neurosciences and fosters interdisciplinary training and research efforts toward meeting these challenges. Graduate Programs in Scientific Computation and Neuroscience are united with the Supercomputing Institute to provide a new paradigm for training graduate students interested in the physical, chemical, and computational sciences. This lowers the barriers to interdisciplinary research, provides opportunities for neuroscientists to pose problems to the quantitative sciences, and provides a catalyst for the cross-fertilization of the two disciplines. Return to Top

Other Collaborators In addition, the Supercomputing Institute collaborates with many other centers of the University of Minnesota. The Supercomputing Institute has cosponsored various research projects, symposia, and workshops with the Aerospace Engineering and Mechanics Department, Army High-Performance Computing Research Center, Institute of Medical Biotechnology, Biological Process Technology Institute, Brain Sciences Center, Center for Interfacial Engineering, Chemical Engineering and Materials Science Department, Computer Science and Engineering Department, Corrosion Research Center, Engineering Research Center for Plasma-Aided Manufacturing, Center for Transportation Studies, Institute of Human Genetics, Laboratory for Computational Science and Engineering, Institute for Mathematics and Its Applications, Limnological Research Center, Medicinal Chemistry Graduate Program, Particle Technology Laboratory, Advanced Bioscience Computing Center, Office of Information Technology, St. Anthony Falls Laboratory, Structural Biology Program, Scientific Computation Graduate Program, and the Theoretical Physics Institute. The Supercomputing Institute is also playing a central role in the University of Minnesota's digital technology initiative. This initiative includes technologies based on computers, electronics, and telecommunications. Key to this initiative is the renovation of Walter Library on the University's East Bank Campus into a Digital Technology Center. This renovation, which has been funded for $53.6 million, will keep the library's historic 1920s decor intact while revamping the interior to include state-of-the-art technology. The Supercomputing Institute will be one of the principal tenants of the Digital Technology Center. Completion of the renovation is scheduled for January 2001. An important part of the Ditigal Technology Initiative is the hiring of four senior faculty in computational biological sciences. Return to Top

Other Programs Supercomputing Institute Research Scholars In addition to providing state-of-the-art supercomputing resources to the University of Minnesota research community, the Supercomputing Institute offers a Research Scholarship Program which provides grants to enhance the supercomputing research programs of University of Minnesota faculty. These grants, which are peer reviewed and competitively awarded, are for the support of research associates who work closely with Supercomputing Institute principal investigators on their research projects. Over the past ten years, 170 Supercomputing Institute Research Scholarships have been awarded. These Research Scholarships have provided an important opportunity for the creation and pursuit of research projects that might not have otherwise been attempted. Supercomputing Institute Undergraduate Interns The Supercomputing Institute's Undergraduate Internship Program promotes undergraduate involvement in ongoing and new research in many fields and provides students with an opportunity to work full-time on challenging and computationally intensive problems in an academic research environment. During ten-week appointments, interns participate in Institute sponsored tutorials specific to high-performance computing and individual laboratory tours led by faculty members. To conclude the program, interns present talks that allow them to share their work with other researchers and to gain experience making scientific presentations. The program allows the students to perform research in close collaboration with faculty investigators and their research groups and to discuss research with faculty members, post-doctoral associates, graduate students, and other interns with similar interests. The program has sponsored 340 interns in its ten years of existence. Return to Top

[Home | 1999 Annual Report | Department Index | Principal Investigator Index]