Wednesday, July 12

MS38
Rounding Error Analysis

10:30 AM-12:30 PM
Rio Mar 3

John von Neumann identified rounding error as a prime source of difficulty in scientific computing. Understanding the effects of rounding error has become an integral part of the development and analysis of numerical algorithms. That work provides an unifying core of concepts and methodologies for diverse numerical fields. Algorithms for new applications and new architectures, as well as the need to solve ever-larger problems, are a continuing source of challenges. This minisymposium surveys recent results, both placing them in an historical context and providing examples of the types of analyses and problems currently being investigated.

10:30-10:55 A Fundamental Theorem for Minimal Data Perturbations

Joseph F. Grcar. Organizer

11:00-11:25 Recent Results in Error Analysis for Linear Equations and Least Squares Problems

Nicholas J. Higham, Organizer

11:30-11:55 A Fast Backward Stable Algorithm for Certain Integral Equations

Shivkumar Chandrasekaran and M. Gu, University of California, Santa Barbara, USA

12:00-12:25 Stability of Structured Hamiltonian Eigensolvers

Francoise Tisseur, University of Manchester, United Kingdom