Research at the Los Alamos Computer Science Institute

Ken Kennedy, David Applegate, Bob Bixby, Alan Carle, Alan Cox, Bill Cook, Keith Cooper, John Dennis, Nate Dean, Rob Fowler, Richard Hanson, Matthias Heinkenschloss, Charlie Hu, Guohua Jin, Petr Kloucek, John Mellor-Crummey, Dan Sorensen, Bill Symes, Frank Toffoletto, Richard Tapia, Linda Torczon, Joe Warren, Yin Zhang, and Willy Zwaenepoel, Rice University; Lennart Johnsson, Barbara Chapman, Jaspal Subhlok, University of Houston; Dan Reed, Ruth Aydt, University of Illinois at Urbana-Champaign; and Jack Dongarra and Antoine Petitet, University of Tennessee at Knoxville

Researchers from two CRPC sites and two affiliated sites at academic institutions are collaborating with Los Alamos National Laboratory (LANL) to establish the Los Alamos Computer Science Institute (LACSI). LACSI was formed to foster computer science and computational science research efforts that are internationally recognized and relevant to LANL.

LACSI was founded with the following goals:

  • To build a presence in computer science research at LANL that is commensurate with the strength of its physics community.
  • To achieve a level of prestige in the computer science community that is on par with the best computer science departments in the nation.
  • To pursue computer science research that is relevant to the goals of high-performance computing programs at LANL.
  • To ensure that there remains a strong focus on high-performance computing in the academic computer science community.

LACSI research is focused on three primary areas, each with several subgroups:

  • Compilation, Systems, and Performance Evaluation of Large Scale Parallel Machines (John Mellor-Crummey, PI)
  • Compiliation Strategies
  • Compilation Issues
  • Tools
  • OpenMP

  • Component Architectures for Distributed Parallel Computing and Rapid Application Development and Composition (Ken Kennedy, PI)
  • Component Architectures for Problem-Solving Environments
  • Application Development Support for Distributed Computing
  • Numerical Component Libraries

  • Computational Mathematics (Yin Zhang, PI)
  • Large-Scale Nonlinear Optimization Algorithms and Software
  • Parallel Tools for the Analysis of Linked Subsystems
  • Code-Based Sensitivity Analysis Project
  • Eigenvalue Methods and Software for ASCI-MPP Systems
  • Integer/Mixed Integer Programming and Large-Scale Problems
  • Methods and Tools for the Solution of Non-Smooth, Multi-Scale, Coupled Models
  • Methods and Software for Inverse and Control Problems

LASCI has major components on site at LANL and Rice University, the lead institution, with participation from the Rice Center for Research on High Performance Software (HiPerSoft), the University of Houston, the University of Illinois at Urbana-Champagne, and the University of Tennessee at Knoxville. LASCI will be a long-term, collaborative research relationship. For more information, contact Danny Powell at danny@rice.edu.

Other Issues of PCR Back to PCR CRPC Home Page