MRS Meetings and Events


EL20.09.04 2023 MRS Fall Meeting

Nanowires as Building Blocks for Unconventional Computing Nanoarchitectures

When and Where

Nov 30, 2023
10:45am - 11:15am

Hynes, Level 3, Room 301



Gianluca Milano1,Carlo Ricciardi2

Istituto Nazionale di Ricerca Metrologica (INRiM)1,Politecnico di Torino2


Gianluca Milano1,Carlo Ricciardi2

Istituto Nazionale di Ricerca Metrologica (INRiM)1,Politecnico di Torino2
G. Milano<sup>1</sup>*, C. Ricciardi<sup>2</sup><br/> <br/><sup>1</sup><i>Advanced Materials Metrology and Life Science Division, INRiM (Istituto Nazionale di Ricerca</i><br/><i>Metrologica), Strada delle Cacce 91, Italy.</i><br/><sup>2</sup><i>Department of Applied Science and Technology, Politecnico di Torino, Corso Duca Degli Abruzzi, 24, Italy.</i><br/> <br/><i>Email: </i><i>[email protected]</i><br/> <br/>Self-assembled computing nanoarchitectures based on interacting nanoobjects have been recently proposed as biologically plausible substrates for neuromorphic computing [1,2]. In these nanonetworks, functionalities arise from the emergent behavior of the system that rely on collective phenomena. Here, we report on NW networks as versatile substrate for physical (<i>in materia</i>) computing. We show that the emergent behavior in these complex networks arise from resistive switching phenomena in single network elements, as revealed by investigating switching properties of single NW and single NW junctions. The emergent spatio-temporal dynamics of these nanoarchitectures in multiterminal characterization is shown that can be exploited for neuromorphic-type of data processing and represent versatile substrates for the implementation of unconventional computing. In the framework of physical reservoir computing (RC), we report on the computing capabilities of these networks by considering different computing tasks such as pattern recognition, time-series prediction and speech recognition. A combined experimental and modeling approach will be discussed, and a theoretical framework based on graph theory to describe complex nanonetworks will be discussed. Going beyond the concept of two terminal measurements, we report on experimental evidence of memory engrams (or memory traces) in NW networks, i.e., physicochemical changes in biological neural networks supposed to endow the representation of experience stored in our brain. Results show that these networks represent a versatile substrate for computing, allowing both encoding and consolidation of information in the same physical substrate, paving the way for a development of new <i>in materia</i> computing paradigms.<br/> <br/>[1] G. Milano, G. Pedretti, M. Fretto, L. Boarino, F. Benfenati, D. Ielmini, I. Valov, C. Ricciardi, Adv. Intell. Syst. <b>2020</b>, 2000096.<br/>[2] G. Milano, G. Pedretti, K. Montano, S. Ricci, S. Hashemkhani, L. Boarino, D. Ielmini, C. Ricciardi, Nat. Mater. <b>2021</b>, DOI 10.1038/s41563-021-01099-9.

Symposium Organizers

Gina Adam, George Washington University
Sayani Majumdar, Tampere University
Radu Sporea, University of Surrey
Yiyang Li, University of Michigan

Symposium Support

APL Machine Learning | AIP Publishing

Publishing Alliance

MRS publishes with Springer Nature