NERSC Staff, Researchers Win Top Computing Awards

Two NERSC staff members and two NERSC clients were honored by the scientific computing community as recipients of top awards at SC98 in Orlando, Florida.

A team of researchers working on the DOE Grand Challenge project Materials, Methods, Microstructure, and Magnetism won the 1998 Gordon Bell Prize for best performance of a supercomputing application. The team won the award for their modeling of 1,024 atoms of a metallic magnet.

The model was run on progressively more powerful Cray T3Es, starting with NERSC's 512-processor machine, and won the prize with a top performance of 657 gigaflops on a 1024-processor T3E. Shortly before the prize was awarded, the group was given access to a 1480-processor machine on Cray's manufacturing center floor and achieved 1.02 teraflops performance--the first time a complete application broke the teraflops barrier.

The prize-winning group included Malcolm Stocks, B. Ujfalussy, Xin-Dong Wang, Xiaoguang Zhang, D. M. C. Nicholson, and W. A. Shelton of Oak Ridge National Laboratory; NERSC staff physicist Andrew Canning; Yang Wang of the Pittsburgh Supercomputing Center; and B. L. Gyorffy of the University of Bristol, UK.

Andrew Canning's role included testing the code, working with Cray to fine-tune it, obtaining access to the fastest computers, and running the code at record-breaking speeds.

Phil Colella, leader of NERSC's Applied Numerical Algorithms Group, received the 1998 Sidney Fernbach Award "for fundamental contributions in the development of software methodologies used to solve numerical partial differential equations, and their application to substantially expand our understanding of shock physics and other fluid dynamics problems."

Andrew Canning	Phil Colella

Phil's award nomination cited his contributions in high-resolution finite difference methods, adaptive mesh refinement, volume-of-fluid methods for fronts and irregular geometries, and multidimensional shock dynamics.

Another 1998 Gordon Bell prize recognized scientists who achieve the best price/performance level on a computer system. The winning team in this category included Greg Kilcup, a physicist at The Ohio State University. And the SC98 award for Best Applications Paper and Best Overall Paper went to a paper co-authored by Robert Harrison of Pacific Northwest National Laboratory. Kilcup and Harrison are both DOE Grand Challenge researchers and NERSC clients.

Earlier, at the Fifth Copper Mountain Conference on Iterative Methods, Mark Adams from the University of California, Berkeley (a student of UCB/NERSC computer scientist James Demmel) won the best student paper competition for his work on a parallel multigrid solver for unstructured finite element method meshes in solid mechanics.