Numerical simulations of neuromorphic behavior in Mott memristive devices

The information age we live in is supported on a physical under-layer of electronic hardware, which originates in condensed matter physics research. The mighty progress made in silicon based technology seemed endless. However, with the smallest feature size of transistors reaching down to mere 5 nm and the large power requirements,

this technology is reaching an unavoidable physical limit. This calls for exploration of new alternatives.

Neuromorphic inspired systems are making fast progress. But they is based either on hardware made with conventional CMOS electronics, or in software, such Deep Neural Networks algorithms that run in neuromorphic chips, which are optimized conventional computers.

Resistive switching phenomena, also known as memristive behavior, opens the way to explore a technological disruptive solution [1]. Namely, to implement simple devices with the required neuromorphic functionalities of neurons and synapses, which would allow to directly build in hardware neural networks systems for artificial intelligence. A an exciting development was to realize that Mott quantum materials can be used to implement artificial spiking neurons. However, the understanding of Mott insulators driven out of equilibrium remains a significant challenge [2]. In this talk we shall describe recent efforts towards making artificial neurons using Mott strongly correlated systems and modeling their facsinating emergent behavior [3].

[1] Challenges in materials and devices for resistive-switching-based neuromorphic computing;

Javier del Valle, Juan Gabriel Ramírez, Marcelo J. Rozenberg, and Ivan K. Schuller; Journal of Applied Physics 124, 211101 (2018)

[2] Subthreshold firing in Mott nanodevices

J. Del Valle, et al.; Nature [01 May 2019, 569(7756):388-392]

[3] Exponential Escape Rate of Filamentary Incubation in Mott Spiking Neurons

Rodolfo Rocco, Javier del Valle, Henry Navarro, Pavel Salev, Ivan K. Schuller, and Marcelo Rozenberg

Phys. Rev. Applied 17, 024028 (2022)

Editors’ Suggestion