RRAM Crossbar-Based Fault-Tolerant Binary Neural Networks (BNNs)

RRAM Crossbar-Based Fault-Tolerant Binary Neural Networks (BNNs)

Gebregiorgis, A.B. (TU Delft Computer Engineering)
Zografou, Artemis (Student TU Delft)
Hamdioui, S. (TU Delft Quantum & Computer Engineering)

Quantum & Computer Engineering

2022

Computation-In Memory (CIM) using RRAM crossbar array is a promising solution to realize energy-efficient neuromorphic hardware, such as Binary Neural Networks (BNNs). However, RRAM faults restrict the applicability of CIM for BNN implementation. To address this issue, we propose a fault tolerance framework to mitigate the impact of RRAM faults on the accuracy of CIM-based BNN hardware. Evaluation results using MNIST, Fashion-MNIST and CIFAR-10 datasets demonstrate that the proposed framework outperforms the related works as it restores more than 99% of the RRAM fault induced accuracy reduction with relatively less overhead.

CIM
fault tolerance
RRAM
BNN

http://resolver.tudelft.nl/uuid:b3f526c0-6453-4222-8a96-7280e9dde55a

https://doi.org/10.1109/ETS54262.2022.9810414

IEEE, Danvers

2023-07-01

978-1-6654-6707-0

Proceedings of the 2022 IEEE European Test Symposium (ETS)

2022 IEEE European Test Symposium (ETS), 2022-05-23 → 2022-05-27, Barcelona, Spain

Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

Institutional Repository

conference paper

© 2022 A.B. Gebregiorgis, Artemis Zografou, S. Hamdioui