Paper Info
Title
Numerical Defect Correction as an Algorithm-Based Fault Tolerance Technique for Iterative Solvers
Abstract
As hardware devices like processor cores and memory sub-systems based on nano-scale technology nodes become more unreliable, the need for fault tolerant numerical computing engines, as used in many critical applications with long computation/mission times, is becoming pronounced. In this paper, we present an Algorithm-based Fault Tolerance (ABFT) scheme for an iterative linear solver engine based on the Conjugated Gradient method (CG) by taking the advantage of numerical defect correction. This method is "pay as you go", meaning that there is practically only a runtime overhead if errors occur and a correction is performed. Our experimental comparison with software-based Triple Modular Redundancy (TMR) clearly shows the runtime benefit of the proposed approach, good fault tolerance and no occurrence of silent data corruption.
Year
DOI
Venue
2011
10.1109/PRDC.2011.26
PRDC
Keywords
Field
DocType
runtime benefit,experimental comparison,algorithm-based fault tolerance,fault tolerant numerical computing,numerical defect correction,good fault tolerance,iterative solvers,hardware device,conjugated gradient method,critical application,runtime overhead,algorithm-based fault tolerance technique,iterative methods,conjugate gradient method,vectors,redundancy,hardware,fault tolerant system,convergence,fault tolerance,triple modular redundancy,fault tolerant,software fault tolerance,conjugate gradient
Gradient method,Computer science,Iterative method,Parallel computing,Algorithm,Triple modular redundancy,Software fault tolerance,Real-time computing,Redundancy (engineering),Fault tolerance,Multi-core processor,Computation
Conference
Citations 
PageRank 
References 
5
0.49
11
Authors
5
Name
Order
Citations
PageRank
Fabian Oboril128826.71
Mehdi B. Tahoori21537163.44
Vincent Heuveline317930.51
Dimitar Lukarski4395.39
Jan-Philipp Weiss5465.37