{"id":10974,"date":"2024-09-15T10:07:15","date_gmt":"2024-09-15T08:07:15","guid":{"rendered":"https:\/\/fermi.univ-tlse3.fr\/?post_type=tribe_events&p=10974"},"modified":"2024-09-16T15:53:12","modified_gmt":"2024-09-16T13:53:12","slug":"brain-like-computation-with-percolating-networks-of-nanoparticles-simon-brown-lpcno-seminar-19-09-2024","status":"publish","type":"tribe_events","link":"https:\/\/fermi.univ-tlse3.fr\/fr\/home\/event\/brain-like-computation-with-percolating-networks-of-nanoparticles-simon-brown-lpcno-seminar-19-09-2024\/","title":{"rendered":"Brain-like Computation with Percolating Networks of Nanoparticles. – (Simon Brown \/ LPCNO \/ Seminar). – 19\/09\/2024"},"content":{"rendered":"

S\u00e9minaire LPCNO<\/strong> <\/span><\/p>\n

Simon Brown, <\/strong><\/span>University of Canterbury, Christchurch, NZ<\/span><\/p>\n

S<\/strong>Seminar room, INSA, Build. 27, Room 2.20 (2nd floor left)<\/span><\/p>\n

Abstract<\/strong> <\/span>
\nSelf-assembled networks of nanoparticles have recently emerged as important candidate systems for brain-like (or neuromorphic) information processing.[1] The essence of the approach is to take advantage of the intrinsic dynamical properties of these networks to implement brain-inspired approaches to computation.[2]<\/p>\n

Our percolating networks of nanoparticles (PNNs, Fig 1(a)) are self-assembled via simple deposition processes that are completely CMOS compatible, making them attractive for integration.[3] The tunnel gaps between particles (Fig 1(b)) turn out to have neuron-like properties, which means that PNNs can be viewed as networks of neurons.[4] Neuron-like electrical spikes are generated when<\/p>\n

We have explored brain-like computation with PNNs in two regimes, beginning with simulations[5,6,7] that allow us to understand the processes and refine parameters, and then moving to experimental demonstrations[8]. At low voltages, the devices are amenable to reservoir computation and we have successfully demonstrated time series prediction, non-linear transformation and spoken digit recognition.[5,8] In the high voltage regime, the spiking behaviour of the \u2018neurons\u2019 has been exploited to perform Boolean logic and MNIST classification[6], and, most recently, optimization tasks including specifically integer factorisation[7]<\/p>\n

[1] J. B. Mallinson et al, Science Advances 5, eaaw8438 (2019).
\n[2] S. Shirai et al, Network Neuroscience 4, 432 (2020).
\n[3] A. Sattar et al, Phys. Rev. Lett. 111, 136808 (2013).
\n[4] R. K. Daniels, et al Neural Networks 154, 122 (2022), Phys. Rev. Applied 20, 034021 (2023).
\n[5] J. B. Mallinson et al, Nanoscale 15, 9663 (2023).
\n[6] S. J. Studholme et al, Nano. Lett. 23, 10594 (2023).
\n[7] S. J. Studholme et al, in press, ACS Nano (2024).
\n[8] J. B. Mallinson et al, Advanced Materials (2024); adma.202402319<\/p>\n

\n

<\/div>