Fig. 3: Collective switching dynamics under DC bias.
From: Avalanches and edge-of-chaos learning in neuromorphic nanowire networks
Simulations using a 100 nanowire, 261 junction NWN at voltage V* = V/Vth = 1.1. a Conductance states of each junction, Gjn, along shortest source-drain path (in units of conductance quantum G0). Junctions are numbered sequentially according to distance from source (#1 closest, #9 farthest). Network conductance, \({G}_{{{\rm{nw}}}}^{* }(t)\) (normalised by source-drain path length), is also shown. b Corresponding junction voltage, \({V}_{{{\rm{jn}}}}^{* }={V}_{{{\rm{jn}}}}/{V}_{{{\rm{set}}}}\), dynamical re-distribution. c–e Snapshots of Gjn distribution (edges shown on a logarithmic colourbar scale) at different time-points during network activation. Node colour indicates voltage on nanowires between source (green asterisk) and drain (red asterisk).
