Memristor Overwrite Logic (MOL) for In-Memory DNN - Université Bretagne Sud Accéder directement au contenu
Communication Dans Un Congrès Année : 2020

Memristor Overwrite Logic (MOL) for In-Memory DNN

Résumé

In-memory computing is a promising solution to address the memory wall challenges in future processing systems. Substantial improvement in performance and energy efficiency is expected, in particular for data intensive applications. A typical use case is neural network applications, where large amount of data should be processed and moved between memory and processing cores. Although several recent works tried to accelerate processing through dedicated parallel hardware designs, data movement cost is still a critical technical challenge. In this context, we propose a novel programmable architecture design for in-memory deep neural networks (DNN) computation. Based on a new logic design style, namely Memristor Overwrite Logic (MOL), specialized computational memory is designed. The original architecture of the proposed computational memory allows to execute multiply-accumulate operations between stored words using MOL. Outstanding features are demonstrated with respect to other recent logic design styles based on emerging non-volatile memory technologies.
Fichier principal
Vignette du fichier
MOL-based DNN.pdf (2.24 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02513024 , version 1 (20-03-2020)

Identifiants

Citer

Khaled Alhaj Ali, Mostafa Rizk, Amer Baghdadi, Jean-Philippe Diguet, Jalal Jomaah. Memristor Overwrite Logic (MOL) for In-Memory DNN. ISCAS 2020 : IEEE International Symposium on Circuits and Systems, Oct 2020, Seville, Spain. ⟨10.1109/ISCAS45731.2020.9180549⟩. ⟨hal-02513024⟩
200 Consultations
143 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More