A new technique is presented that allows to execute the preconditioned conjugate gradient (PCG) solver on approximate hardware while ensuring correct solver results. This technique expands the scope of approximate computing to scientific and engineering applications. The changing error resilience of PCG during the solving process is exploited by different levels of approximation which trade off numerical accuracy and hardware utilization. Such approximation levels are determined at runtime by periodically estimating the error resilience. An efficient fault tolerance technique allows reductions in hardware utilization by ensuring the continued exploitation of maximum allowed energy-accuracy trade-offs. Experimental results show that the hardware utilization is reduced on average by 14.5% and by up to 41.0% compared to executing PCG on accurate hardware.
%0 Conference Paper
%1 SchoeBW2016
%A Schöll, Alexander
%A Braun, Claus
%A Wunderlich, Hans-Joachim
%B Proceedings of the IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems (DFT'16)
%D 2016
%K AxC CCG PCG SimTech approximate computing conjugate error-correction error-detection fault-tolerance gradient linear myown preconditioned solver sparse systems
%P 21-26
%R http://dx.doi.org/10.1109/DFT.2016.7684063
%T Applying Efficient Fault Tolerance to Enable the Preconditioned Conjugate Gradient Solver on Approximate Computing Hardware
%X A new technique is presented that allows to execute the preconditioned conjugate gradient (PCG) solver on approximate hardware while ensuring correct solver results. This technique expands the scope of approximate computing to scientific and engineering applications. The changing error resilience of PCG during the solving process is exploited by different levels of approximation which trade off numerical accuracy and hardware utilization. Such approximation levels are determined at runtime by periodically estimating the error resilience. An efficient fault tolerance technique allows reductions in hardware utilization by ensuring the continued exploitation of maximum allowed energy-accuracy trade-offs. Experimental results show that the hardware utilization is reduced on average by 14.5% and by up to 41.0% compared to executing PCG on accurate hardware.
@inproceedings{SchoeBW2016,
abstract = {A new technique is presented that allows to execute the preconditioned conjugate gradient (PCG) solver on approximate hardware while ensuring correct solver results. This technique expands the scope of approximate computing to scientific and engineering applications. The changing error resilience of PCG during the solving process is exploited by different levels of approximation which trade off numerical accuracy and hardware utilization. Such approximation levels are determined at runtime by periodically estimating the error resilience. An efficient fault tolerance technique allows reductions in hardware utilization by ensuring the continued exploitation of maximum allowed energy-accuracy trade-offs. Experimental results show that the hardware utilization is reduced on average by 14.5% and by up to 41.0% compared to executing PCG on accurate hardware.},
added-at = {2018-03-19T16:15:07.000+0100},
author = {Schöll, Alexander and Braun, Claus and Wunderlich, Hans-Joachim},
biburl = {https://puma.ub.uni-stuttgart.de/bibtex/28b341984a107175be05eebdda39e6c12/clausbraun},
booktitle = {Proceedings of the IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems (DFT'16)},
doi = {http://dx.doi.org/10.1109/DFT.2016.7684063},
file = {http://www.iti.uni-stuttgart.de/fileadmin/rami/files/publications/2016/DFT_SchoeBW2016.pdf},
interhash = {819e882fc0ec03e0c6e332411bfbf42d},
intrahash = {8b341984a107175be05eebdda39e6c12},
keywords = {AxC CCG PCG SimTech approximate computing conjugate error-correction error-detection fault-tolerance gradient linear myown preconditioned solver sparse systems},
pages = {21-26},
timestamp = {2018-03-19T15:34:51.000+0100},
title = {{Applying Efficient Fault Tolerance to Enable the Preconditioned Conjugate Gradient Solver on Approximate Computing Hardware}},
year = 2016
}