Experimental demonstration and tolerancing of a large-scale neural network (165,000 synapses), using phase-change memory as the synaptic weight element

By Geoffrey W. Burr, R. Shelby, Severin Sidler, C. D. Nolfo, Junwoo Jang, I. Boybat, Rohit S. Shenoy, Pritish Narayanan, K. Virwani, E. Giacometti and 2 more. 2014 IEEE International Electron Devices Meeting, 2015

724 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Electrical and Electronic Engineering

View Paper: https://www.semanticscholar.org/paper/89d6ab4ac6aac98209173d8cda76a31ff8687119

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Electrical and Electronic Engineering

Citations in the corpus

A Survey of Neuromorphic Computing and Neural Networks in Hardware (2017, methodology): “[1906], [1907], [1922], [2080], [2086], [2100]–[2103], [2115],”
In-memory computing on a photonic platform (2018, neutral): “, for forward and backward propagation in deep neural networks (23, 24)].”
ISAAC: A Convolutional Neural Network Accelerator with In-Situ Analog Arithmetic in Crossbars (2016, methodology): “Bojnordi et al. design an accelerator for Boltzmann machine [4], using in-situ memristor computation to estimate a sum of products of two 1-bit numbers and a 16-bit weight.”
In-memory computing with resistive switching devices (2018, background): “The concept of the accumulating PCM counter can be extended to several applications in the realm of analogue computing, such as the decomposition in prime factors 59 , the gradual potentiation of a PCM artificial synapse 60,61 , and the logic summation, namely digital OR within a single PCM 53 .”
A review of emerging non-volatile memory (NVM) technologies and applications (2016, background): “The impact of imperfection of NVM devices on the performance of synaptic network is an important research topic [124].”
Towards Oxide Electronics: a Roadmap (2019, background): “Oxide memristors will need improvements to overcome their competitors including technologically-mature chalcogenide-based phase change materials [437,438].”
A compute-in-memory chip based on resistive random-access memory (2022, methodology): “Although these studies proposed various techniques to mitigate the impacts of analogue-related hardware non-idealities on inference accuracy, the AI benchmark results reported were often obtained by performing software emulation based on characterized device data 3,5,21,24 .”