Search

Atomic Switch Networks (ASN) comprising silver iodide (AgI) junctions, a material previously unexplored as functional memristive elements within highly-interconnected nanowire networks, were employed as a neuromorphic substrate for physical Reservoir Computing (RC). This new class of ASN-based devices has been physically characterized and utilized to classify spoken digit audio data, demonstrating the utility of substrate-based device architectures where intrinsic material properties can be exploited to perform computation in-materio. This work demonstrates high accuracy in the classification of temporally analyzed Free-Spoken Digit Data (FSDD). These results expand upon the class of viable memristive materials available for the production of functional nanowire networks and bolster the utility of ASN-based devices as unique hardware platforms for neuromorphic computing applications involving memory, adaptation and learning., 11 pages, 7 figures

Artificial realizations of the mammalian brain alongside their integration into electronic components are explored through neuromorphic architectures, neuroarchitectectonics, on CMOS compatible platforms. Exploration of neuromorphic technologies continue to develop as an alternative computational paradigm as both capacity and capability reach their fundamental limits with the end of the transistor-driven industrial phenomenon of Moore’s law. Here, we consider the electronic landscape within neuromorphic technologies and the role of the atomic switch as a model device. We report the fabrication of an atomic switch network (ASN) showing critical dynamics and harness criticality to perform benchmark signal classification and Boolean logic tasks. Observed evidence of biomimetic behavior such as synaptic plasticity and fading memory enable the ASN to attain a cognitive capability within the context of artificial neural networks.

Complexity is an increasingly crucial aspect of societal, environmental and biological phenomena. Using a dense unorganized network of synthetic synapses it is shown that a complex adaptive system can be physically created on a microchip built especially for complex problems. These neuro-inspired atomic switch networks (ASNs) are a dynamic system with inherent and distributed memory, recurrent pathways, and up to a billion interacting elements. We demonstrate key parameters describing self-organized behavior such as non-linearity, power law dynamics, and multistate switching regimes. Device dynamics are then investigated using a feedback loop which provides control over current and voltage power-law behavior. Wide ranging prospective applications include understanding and eventually predicting future events that display complex emergent behavior in the critical regime.

Developments in computing hardware are constrained by the operating principles of complementary metal oxide semiconductor (CMOS) technology, fabrication limits of nanometer scaled features, and difficulties in effective utilization of high density interconnects. This set of obstacles has promulgated a search for alternative, energy efficient approaches to computing inspired by natural systems including the mammalian brain. Atomic switch network (ASN) devices are a unique platform specifically developed to overcome these current barriers to realize adaptive neuromorphic technology. ASNs are composed of a massively interconnected network of atomic switches with a density of ∼109 units/cm2 and are structurally reminiscent of the neocortex of the brain. ASNs possess both the intrinsic capabilities of individual memristive switches, such as memory capacity and multi-state switching, and the characteristics of large-scale complex systems, such as power-law dynamics and non-linear transformations of input signals. Here we describe the successful nanoarchitectonic fabrication of next-generation ASN devices using combined top-down and bottom-up processing and experimentally demonstrate their utility as reservoir computing hardware. Leveraging their intrinsic dynamics and transformative input/output (I/O) behavior enabled waveform regression of periodic signals in the absence of embedded algorithms, further supporting the potential utility of ASN technology as a platform for unconventional approaches to computing.