Estimation of error propagation in multiprocessor systems

Authors:
A. Ivanyi;M. M. Ivanyi;P. Ivanyi
Affiliations:
Pollack Mihaly Faculty of Engineering, University of Pecs, Boszorkany u. 2, H-7624 Pecs, Hungary;Pollack Mihaly Faculty of Engineering, University of Pecs, Boszorkany u. 2, H-7624 Pecs, Hungary and UVATERV Ltd., Dombovari u. 17-19, H-1117 Budapest, Hungary;Pollack Mihaly Faculty of Engineering, University of Pecs, Boszorkany u. 2, H-7624 Pecs, Hungary
Venue:
Advances in Engineering Software
Year:
2013

Citing 8
Cited 0

Error estimates for parallel shooting using initial or boundary value methods

NUMDIFF-7 Selected papers of the seventh conference on Numerical treatment of differential equations
Fixed-point error analysis and an efficient array processor design of two-dimensional sliding DFT

Signal Processing
Algorithms for real-time scheduling of error-cumulative tasks based on the imprecise computation approach

Journal of Systems Architecture: the EUROMICRO Journal
Parallel application performance on shared high performance reconfigurable computing resources

Performance Evaluation - Performance modelling and evaluation of high-performance parallel and distributed systems
A parallel algorithm for accurate dot product

Parallel Computing
Scheduling in a dynamic heterogeneous distributed system using estimation error

Journal of Parallel and Distributed Computing
Algorithm-based fault tolerance applied to high performance computing

Journal of Parallel and Distributed Computing
Efficient and accurate computation of upper bounds of approximation errors

Theoretical Computer Science

Quantified Score

Hi-index	0.00

Visualization

Abstract

In high performance computation where the processors evaluate their required algorithm in serial or parallel action the estimation of the error and the error propagation during the computation is important. In the paper the system of the grid-connected processors is considered as a system. To determine the error of the processor a data is sent through on it under its operation and the output data has some error. Taking into account the error of the output data, it can be decomposed into system error and input error. In the paper determination of the error for different processor configurations are evaluated and applied to some grid-connected processor system. Numerical examples represent the developed theory.