Extending Backward Error Assertions to Tolerance of Large Errors in Floating Point Computations

Authors:
Patrick Fitzpatrick
Affiliations:
Univ. College, and National Microelectronics Research Center, Cork, Ireland
Venue:
IEEE Transactions on Computers
Year:
1997

Citing 8
Cited 0

An analysis of algorithm-based fault tolerance techniques

Journal of Parallel and Distributed Computing
Fault-Tolerant Matrix Triangularizations on Systolic Arrays

IEEE Transactions on Computers
A Linear Algebraic Model of Algorithm-Based Fault Tolerance

IEEE Transactions on Computers
On multiple error detection in matrix triangularizations using checksum methods

Journal of Parallel and Distributed Computing
Iterative refinement enhances the stability of QR factorization methods for solving linear equations

BIT
Floating Point Fault Tolerance with Backward Error Assertions

IEEE Transactions on Computers - Special issue on fault-tolerant computing
On fault tolerant matrix decomposition

Journal of VLSI Signal Processing Systems - Special issue on the Canadian conference on VLSI
A New Error Analysis Based Method for Tolerance Computation for Algorithm-Based Checks

IEEE Transactions on Computers

Quantified Score

Hi-index	14.98

Visualization

Abstract

The use of backward error assertions combined with iterative refinement has been suggested for the correction of small fault induced errors in the floating point solution of linear systems. We extend this to the correction of large errors, typically caused by the failure of a single processor (or column of processors) in an array.