Thursday, March 13

MS20
Parallel Combinatorial Scientific Computing - Part I of II

10:00 AM - 12:00 PM
Room: Roswell

For Part II, see MS36

Combinatorial algorithms play a crucial enabling role in computational science, and as problems increase in size and complexity, this role continues to grow. Although researchers have been working on these types of problems for decades it is only in the last five years that the community has organized under the banner of Combinatorial Scientific Computing (CSC). One example of the increased focus generated by CSC is the establishment of the SciDAC institute CSCAPES for addressing CSC and petascale simulations.

Since a number of the applications that motivate CSC-type problems are typically run on parallel computers, it is only natural that the CSC solutions should also be parallel; otherwise, the combinatorial parts might become the bottlenecks. The current mini-symposium aims at displaying the state of the art in applying parallel methods to CSC-type problems, such as parallel preconditioners, computing matchings and colorings in parallel, and parallel graph partitioning.

Organizer: Rob Bisseling
Utrecht University, The Netherlands
Fredrik Manne
University of Bergen, Norway

10:00-10:25 Parallel Issues in Combinatorial Scientific Computing

Rob Bisseling, Utrecht University, The Netherlands

10:30-10:55 A Scalable Algorithmic Framework for Parallel Graph Coloring on Distributed Memory Computers

Assefaw H. Gebremedhin, Purdue University; Erik G. Boman, Sandia National Laboratories; Doruk Bozdag, Ohio State University; Umit V. Catalyurek, The Ohio State University; Fredrik Manne, University of Bergen, Norway

11:00-11:25 Parallel Partitioning Techniques in Circuit Simulation

Robert J. Hoekstra and Heidi K. Thornquist, Sandia National Laboratories

11:30-11:55 Efficient and Scalable Parallel Graph Partitioning Methods

Francois Pellegrini and Jun-Ho Her, INRIA Futurs - LaBRI, France