Authors: Enrico Prati
Addresses: Istituto di Fotonica e Nanotecnologie, Consiglio Nazionale delle Ricerche, Piazza Leonardo da Vinci 32, I-20133 Milano, Italy
Abstract: This paper analyses the advancements of atomic scale nanoelectronics towards quantum neuromorphics. First, the key properties of elementary combinations of a few neurons, namely long- and short-term plasticity, spike-timing dependent plasticity (associative plasticity), quantumness and stochastic effects, and their potential computational employment are summarised. Next, several atomic scale device technologies are developed to control electron transport at the atomic level, including single atom implantation for atomic arrays and CMOS quantum dots, single atom memories, Ag2S and Cu2S atomic switches, hafnium-based RRAMs, organic material based transistors, and Ge2Sb2Te5 (GST) synapses. Each material/method was proved successful in achieving some of the properties observed in real neurons. This paper compares the different methods towards the creation of a new generation of naturally inspired and biophysically meaningful artificial neurons, in order to replace the rigid CMOS based neuromorphic hardware. The most challenging aspect to address appears to be to obtain both the stochastic/quantum behaviour and the associative plasticity, which are currently observed only below and above 20 nm length scale respectively, by employing the same material.
Keywords: artificial neurons; artificial synapses; atomic scale nanoelectronics; quantum neuromorphics; plasticity; nanotechnology; electron transport; single atom implantation; atomic arrays; CMOS quantum dots; single atom memories; atomic switches; hafnium-based RRAMs; organic materials; transistors; stochastic behaviour; quantum behaviour.
International Journal of Nanotechnology, 2016 Vol.13 No.7, pp.509 - 523
Published online: 17 Aug 2016 *Full-text access for editors Access for subscribers Purchase this article Comment on this article