Science-Driven System Architecture Team Staff

John Shalf's background is in electrical engineering: he spent time in graduate school at Virginia Tech working on a C-compiler for the SPLASH-2 FPGA-based computing system, and at Spatial Positioning Systems Inc. (now ArcSecond) he worked on embedded computer systems. John first got started in HPC at the National Center for Supercomputing Applications (NCSA) in 1994, where he provided software engineering support for a number of scientific applications groups. While working for the General Relativity Group at the Albert Einstein Institute in Potsdam Germany, he helped develop the first implementation of the Cactus Computational Toolkit, which is used for numerical solutions to Einstein's equations for General Relativity and which enables modeling of black holes, neutron stars, and boson stars. John joined Berkeley Lab in 2000 and has worked in the Visualization Group, on the RAGE robot, and on various projects in the Future Technologies Group.

Andrew Canning works on the programming and algorithmic developments necessary to run codes on parallel machines, specializing in materials science applications. Along with a team of colloborating scientists at Oak Ridge National Lab, Pittsburgh Supercomputing Center, and the University of Bristol (UK), Andrew won the 1998 Gordon Bell Prize for the fastest parallel application, which modeled 1,024 atoms of a metallic magnet. Although the team won for their 657 Gigaflop/s performance level, they subsequently were able to run the application at more than one Teraflop/s. Andrew has a B.S. in theoretical physics and astronomy from the University of Glasgow and a Ph.D. in statistical physics from the University of Edinburgh. For three years he was an employee of Cray Research in Lausanne, Switzerland, developing parallel codes and algorithms for materials science applications on the Cray T3D parallel computer.

Esmond Ng, Group Leader of the Scientific Computing Group, has been involved in the development and implementation of sparse matrix algorithms since 1979, and has been involved in R&D management in scientific computing since 1995. He became one of the co-authors of the well-known sparse matrix package, SPARSPAK, when he was a graduate student at the University of Waterloo. At ORNL, Esmond was one of the first researchers to develop and implement efficient algorithms for sparse matrix computation on parallel computer architectures. He and a colleague, Dr. Barry W. Peyton, worked on sparse matrix algorithms that are specifically designed for computers that have memory hierarchy. Some of the sparse matrix codes they developed have been incorporated into the scientific computing libraries of several computer vendors, as well as in Matlab. Esmond earned his bachelor's, master's, and Ph.D. in Computer Science at the University of Waterloo in Canada.

Leonid Oliker is the point man for computer architectures of the future -- MTA, IRAM, etc. Lenny joined NERSC as a post-doc in the Scientific Computing Group, where he co-authored the "Best Paper of SC99," "Parallelization of a Dynamic Unstructured Application Using Three Leading Paradigms."

Erich Strohmaier, one of the co-founders of the twice-yearly TOP500 listing of the world's most powerful computers, is a specialist in evaluating and optimizing performance of high-performance computing systems. He began working on what would become the TOP500 list, which ranks computers according to how well they run the Linpack benchmark, in 1990 in his native Germany. The list debuted in 1993, and in 1995 Erich moved to Tennessee to work with Jack Dongarra, the team's American representative. Erich was born in Bavaria and earned degrees in physics and theoretical physics at the universities in Heidelberg and Mannheim.

Lin-Wang Wang earned his Ph.D. in Theoretical Solid State Physics at Cornell University. His research interests include large-scale total energy calculations for material simulations, nanoscale electronic structure calculations, alternatives to local-density approximation methods, and software applications.

Before joining Berkeley Lab, Harvey Wasserman spent 24 years working on workload characterization, benchmarking, and system evaluation for the high performance computing program at Los Alamos National Laboratory. He is also active in the Supercomputing conference series where he is executive chair of the Technical Program for SC2007.