Adaptive verifiability-driven strategy for evolutionary approximation of arithmetic circuits

Adaptive verifiability-driven strategy for evolutionary approximation of arithmetic circuits

WoS Article

We present a novel approach for designing complex approximate arithmetic circuits that trade correctness for power consumption and play important role in many energy-aware applications. Our approach integrates in a unique way formal methods providing formal guarantees on the approximation error into an evolutionary circuit optimisation algorithm. The key idea is to employ a novel adaptive search strategy that drives the evolution towards promptly verifiable approximate circuits. As demonstrated in an extensive evaluation including several structurally different arithmetic circuits and target precisions, the search strategy provides superior scalability and versatility with respect to various approximation scenarios. Our approach significantly improves capabilities of the existing methods and paves a way towards an automated design process of provably-correct circuit approximations.

We present a novel approach for designing complex approximate arithmetic circuits that trade correctness for power consumption and play important role in many energy-aware applications. Our approach integrates in a unique way formal methods providing formal guarantees on the approximation error into an evolutionary circuit optimisation algorithm. The key idea is to employ a novel adaptive search strategy that drives the evolution towards promptly verifiable approximate circuits. As demonstrated in an extensive evaluation including several structurally different arithmetic circuits and target precisions, the search strategy provides superior scalability and versatility with respect to various approximation scenarios. Our approach significantly improves capabilities of the existing methods and paves a way towards an automated design process of provably-correct circuit approximations.

Approximate computing, 
Energy efficiency, 
Circuit optimisation, 
Genetic programming

Approximate computing, 
Energy efficiency, 
Circuit optimisation, 
Genetic programming

ČEŠKA, M.; MATYÁŠ, J.; MRÁZEK, V.; SEKANINA, L.; VAŠÍČEK, Z.; VOJNAR, T.

2021

22.06.2020

Applied Soft Computing

1568-4946

APPLIED SOFT COMPUTING

95

106466

Kingdom of the Netherlands

1

17

17

https://www.fit.vut.cz/research/publication/12312/

preprint
main