Research and Development Efforts by PSC Staff

---

Besides providing resources and support to enable U.S. researchers to conduct leading-edge scientific research, PSC staff members perform important research in their own areas of interest.

Publications

PSC staff write Technical Reports in their areas of interest.

The Networking Research Group publishes numerous papers in networking research.

Research in collaboration with the Department of Energy

Federal Energy Technology Center

A partnership between PSC and FETC has produced results in combustion and fossil fuel conversion as well as laying the groundwork for the formation of the Super Computing Science Consortium.

Accelerated Strategic Computing Initiative

Strong-Shock Interface Interactions
Dr. Nick Nystrom of PSC has been working with Dr. Ravi Samtaney of NASA Ames and Dr. Daniel Meiron of CIT on the interactions of shock waves with contact discontinuities, such as an interface separating two gases.

Wake Induced Boundary Layer Transition in Turbomachinery
Dr. Ravi Subramanya of PSC has been collaborating with Prof. Paul Durbin and Dr. Xiaohua Wu of Stanford University on direct numerical simulation (DNS) of wakes impinging on a flat plate to model the transition of the boundry layer from laminar flow to turbulent flow.

Astrophysical Thermonuclear Flashes
Scientists from the Center on Astrophysical Thermonuclear Flashes have been studying the problem of thermonuclear flashes on the surfaces of compact stars such as neutron stars and white dwarfs, and in the interiors of white dwarfs. PSC's John Urbanic has been helping them to expand their studies to three dimensions.

The Relativistic Heavy Ion Collider at Brookhaven National Lab

Simulating Relativistic Heavy Ion Collisions in the STAR Detector
Dr. Nick Nystrom of PSC is working with researchers at Brookhaven National Lab to simulate collision events on the STAR (Solenoidal Tracker At RHIC) detector.

Networking

The PSC Networking Research group is actively involved in developing networking tools and software, authoring networking research papers, and compiling information on network performance and analysis.

NLANR:NCNE^sm - The National Center for Network Engineering, a component of the National Laboratory for Applied Network Research, provides network engineering support services for universities and research sites connected to the high performance research network infrastructure, such as the vBNS, the NSF Connections Program, and the Internet2 project.

ncne.net maintains the NCNE^sm GigaPop, a regional network aggregation point providing high-speed commodity and research network access to sites in Western and Central Pennsylvania.

Biomedical research:

The National Resource for Biomedical Supercomputing pursues leading edge research in high performance computing and the life sciences, and fosters exchange between PSC expertise in computational science and biomedical researchers nationwide.

Key HIV protein research: Dr. Marcela Madrid of the PSC and her collaborators simulated the movement and subdomain rearrangement of reverse transcriptase, an HIV enzyme targeted by AIDS drugs.

High-performance Computing

The Advanced Systems group conducts research into high-performance computing systems, including the development of the CLAN cluster. They are collaborating in many other projects, including the Visible Human Project, the NIH Collaboratory, (SC)², and the HUBS consortium.

Several projects involving Grid computing technology are also underway.

Neural Sciences

Neural Science research includes Functional Magnetic Resonance Imaging research and parallel simulation in Computational Neurobiology.

Scientific Visualization

The Strategic Applications group produces animations and other scientific graphics and also develops graphics software.

Fluid Flow Modeling

PaRSI is a program which simulates rotor-stator interactions. The paper based on this research received the 1997 Westinghouse Science and Technology Center Technical Publication Award.

Materials Science

In 1998, Yang Wang of the PSC collaborated with scientists at Oak Ridge National Laboratory, National Energy Research Supercomputing Center and the University of Bristol, UK, to win the Gordon Bell prize for best acheivement in high-performance computing. Their first-principles simulation of complex magnetic properties is the world's first fully fledged scientific application to sustain more than one Teraflop. This was accomplished on a 1480-processor T3E-1200 system at Cray Research.

Cosmology

The Grand Challenge project in Cosmology brings Grand-Challenge scale computing power to bear on fundamental problems in cosomology.

Heterogeneous Computing:

Latency-Free Metacomputing: Researchers at the PSC and the High Performance Computing Center at Stuttgart University have run an application on two coupled 512 processor Cray T3Es with performance equivalent to a single 1024 processor machine.

Additional information on this project is available in a PSC news release and a slide presentation given at Supercomputing '98.

The Metacomputing Project: PSC, Oak Ridge National Labs, and Sandia National Labs are collaborating to link disparate, highly parallel systems at their respective sites across a high-performance network. These linked machines are used to perform computations too large to be carried out on any single computer.

Applying this work to a large-scale materials science computation earned the collaboration a Gold Medal in the Concurrency category of the High Performance Computing Challenge held at Supercomputing '96. The annual Supercomputing conference was held in Pittsburgh on November 17-22.

High Performance/Parallel Applications:

• Applications Needs for High-Performance Scalable Systems
• Parallel Applications Technology Program (PATP)
• HPCC research at PSC: An Example

Software

• PSC developed software, distributed via anonymous ftp.