Self-organized NW networks can provide a new paradigm for the realization of neuromorphic hardware. The concept of nanoarchitecture, where the mutual interaction among a huge number of nano parts causes new functionalities to emerge, resembles our brain, where an emergent behaviour arises from the synaptic interactions among a huge number of neurons. Besides reservoir computing that represents one of the most promising computing paradigms to be implemented on these nanoarchitectures, unconventional computing frameworks able to process sensor inputs from the environment can be explored for online adapting of robot behavior. In perspective, more complex network dynamics can be explored by realizing computing nanoarchitectures composed of multiple interconnected networks or by stimulating networks with heterogeneous stimuli. In this scenario, NW networks that can learn and adapt when externally stimulated - thus mimicking the processes of experience-dependent synaptic plasticity that shapes connectivity of our nervous system - would not only represent a breakthrough platform for neuro-inspired computing but could also facilitate the understanding of information processing in our brain, where structure and functionalities are intrinsically related.

2022 Roadmap on Neuromorphic Computing and Engineering / Christensen, Dennis Valbjørn; Dittmann, Regina; Linares-Barranco, Bernabe; Sebastian, Abu; Le Gallo, Manuel; Redaelli, Andrea; Slesazeck, Stefan; Mikolajick, Thomas; Spiga, Sabina; Menzel, Stephan; Valov, Ilia; Milano, Gianluca; Ricciardi, Carlo; Liang, Shi-Jun; Miao, Feng; Lanza, Mario; Quill, Tyler J.; Keene, Scott Tom; Salleo, Alberto; Grollier, Julie; Markovic, Danijela; Mizrahi, Alice; Yao, Peng; Yang, J. Joshua; Indiveri, Giacomo; Strachan, John Paul; Datta, Suman; Vianello, Elisa; Valentian, Alexandre; Feldmann, Johannes; Li, Xuan; Pernice, Wolfram HP; Bhaskaran, Harish; Furber, Steve; Neftci, Emre; Scherr, Franz; Maass, Wolfgang; Ramaswamy, Srikanth; Tapson, Jonathan; Panda, Priyadarshini; Kim, Youngeun; Tanaka, Gouhei; Thorpe, Simon; Bartolozzi, Chiara; Cleland, Thomas A; Posch, Christoph; Liu, Shih-Chii; Panuccio, Gabriella; Mahmud, Mufti; Mazumder, Arnab Neelim; Hosseini, Morteza; Mohsenin, Tinoosh; Donati, Elisa; Tolu, Silvia; Galeazzi, Roberto; Christensen, Martin Ejsing; Holm, Sune; Ielmini, Daniele; Pryds, Nini. - In: NEUROMORPHIC COMPUTING AND ENGINEERING. - ISSN 2634-4386. - ELETTRONICO. - (2022). [10.1088/2634-4386/ac4a83]

2022 Roadmap on Neuromorphic Computing and Engineering

Spiga, Sabina;Valov, Ilia;Milano, Gianluca;Ricciardi, Carlo;
2022

Abstract

Self-organized NW networks can provide a new paradigm for the realization of neuromorphic hardware. The concept of nanoarchitecture, where the mutual interaction among a huge number of nano parts causes new functionalities to emerge, resembles our brain, where an emergent behaviour arises from the synaptic interactions among a huge number of neurons. Besides reservoir computing that represents one of the most promising computing paradigms to be implemented on these nanoarchitectures, unconventional computing frameworks able to process sensor inputs from the environment can be explored for online adapting of robot behavior. In perspective, more complex network dynamics can be explored by realizing computing nanoarchitectures composed of multiple interconnected networks or by stimulating networks with heterogeneous stimuli. In this scenario, NW networks that can learn and adapt when externally stimulated - thus mimicking the processes of experience-dependent synaptic plasticity that shapes connectivity of our nervous system - would not only represent a breakthrough platform for neuro-inspired computing but could also facilitate the understanding of information processing in our brain, where structure and functionalities are intrinsically related.
File in questo prodotto:
File Dimensione Formato  
2022_roadmap_NCE.pdf

accesso aperto

Tipologia: 2. Post-print / Author's Accepted Manuscript
Licenza: Creative commons
Dimensione 21.58 MB
Formato Adobe PDF
21.58 MB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

Caricamento pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11583/2959714