Outstanding Bit Error Tolerance of Resistive RAM-Based Binarized Neural Networks - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2019

Outstanding Bit Error Tolerance of Resistive RAM-Based Binarized Neural Networks

Résumé

Resistive random access memories (RRAM) are novel nonvolatile memory technologies, which can be embedded at the core of CMOS, and which could be ideal for the in-memory implementation of deep neural networks. A particularly exciting vision is using them for implementing Binarized Neural Networks (BNNs), a class of deep neural networks with a highly reduced memory footprint. The challenge of resistive memory, however, is that they are prone to device variation, which can lead to bit errors. In this work we show that BNNs can tolerate these bit errors to an outstanding level, through simulations of networks on the MNIST and CIFAR10 tasks. If a standard BNN is used, up to 10 −4 bit error rate can be tolerated with little impact on recognition performance on both MNIST and CIFAR10. We then show that by adapting the training procedure to the fact that the BNN will be operated on error-prone hardware, this tolerance can be extended to a bit error rate of 4 × 10 −2. The requirements for RRAM are therefore a lot less stringent for BNNs than more traditional applications. We show, based on experimental measurements on a RRAM Hf O2 technology, that this result can allow reduce RRAM programming energy by a factor 30.
Fichier principal
Vignette du fichier
Hirtzlin_AICAS2019.pdf (2.39 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02159142 , version 1 (18-06-2019)

Identifiants

Citer

T Hirtzlin, Marc Bocquet, J.-O Klein, E. Nowak, E. Vianello, et al.. Outstanding Bit Error Tolerance of Resistive RAM-Based Binarized Neural Networks. IEEE International Conference on Artificial Intellignence Circuits and Systems (AICAS), Mar 2019, Hsinshu, Taiwan. ⟨10.1109/AICAS.2019.8771544⟩. ⟨hal-02159142⟩
115 Consultations
117 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More