Search

Reservoir computing (RC) is possible using physical systems. We have previously proposed an RC for ideal atomic switches. When temporal current fluctuations (noise) from the measurement of actual atomic switches are introduced into the proposed RC, performance degrades significantly. To address this issue, we propose novel methods for increasing the operating current range and observing the atomic switch several times to determine the average noise. Consequently, the memory capacity of the RC model increased, despite the presence of noise. To improve the precision of RC, we investigated the capacity and showed that changing the time constant of atomic switches results in an improvement.

We propose a simple model for an atomic switch-based decision maker (ASDM), and show that, as long as its total number of metal atoms is conserved when coupled with suitable operations, an atomic switch system provides a sophisticated ``decision-making'' capability that is known to be one of the most important intellectual abilities in human beings. We considered a popular decision-making problem studied in the context of reinforcement learning, the multi-armed bandit problem (MAB); the problem of finding, as accurately and quickly as possible, the most profitable option from a set of options that gives stochastic rewards. These decisions are made as dictated by each volume of precipitated metal atoms, which is moved in a manner similar to the fluctuations of a rigid body in a tug-of-war game. The ``tug-of-war (TOW) dynamics'' of the ASDM exhibits higher efficiency than conventional reinforcement-learning algorithms. We show analytical calculations that validate the statistical reasons for the ASDM to produce such high performance, despite its simplicity. Efficient MAB solvers are useful for many practical applications, because MAB abstracts a variety of decision-making problems in real-world situations where an efficient trial-and-error is required. The proposed scheme will open up a new direction in physics-based analog-computing paradigms, which will include such things as ``intelligent nanodevices'' based on self-judgment.

The invention of the internet began the age of information as well as exponentially increased the number of complex systems in our world. As the age of information comes to an end, so does the persevering trend known as Moore's Law. This means that the number of circuit elements on an integrated chip will no longer double every two years, nor will the processing speed of computers. Personal computers utilize the Von Neumann architecture which separates storage from processing. This separation causes information transfer lags as a computer processes information much faster than it can be fetched from information storage. Thus, to circumvent both the limitations on elemental packing, and areal density a movement into neuromorphic hardware has occurred. Neuromorphic chips seek to emulate brain-like processing of information through low-power, highly parallel, densely interconnected, and closely packed individual elements which have a non-intuitive entangled relationship. This work explored the potential of atomic switch networks (ASNs) for reservoir, natural, and unconventional computing, provides evidence for ASNs as complex adaptive systems operating in and around the edge of chaos, presents a new material for use in ASNs, and evaluates spoken digit recognition using reservoir computing. A second project herein explores the maturation of human pluripotent stem cell-derived cardiomyocytes for use in studying heart disease, which is the leading cause of death in the world. Maturation of these cells is significant to the field. Via a chemically defined differentiation regimen with a monolayer cell culture technique on top of a multi-electrode array for real-time measurements of electrophysiological properties, in vivo development was reproduced. Both systems described above required data science analysis of time-series multi-electrode array information.

10-year journey of the atomic switch in nanospace is finally reaching practical use in space satellites several thousand kilometers above earth, MANA E-BULLETIN. 31 (2017)

The emergent dynamical behaviors of biological neuronal networks and other natural, complex systems point towards new computing paradigms which can overcome limitations of digital computers. This work catalogues the development and characterization of an electronic circuit purpose built to exhibit emergent behaviors intended for use in neuromorphic computation. These circuits, atomic switch networks (ASNs), are fabricated through a self-assembly process that yields a highly interconnected network of silver nanowires with embedded inorganic synapses known as atomic switches. When stimulated with external bias voltage, ASNs are shown to possess the synaptic and memory properties of individual atomic switches, as well as network-specific behavior consisting of distributed, system wide switching events. These emergent behaviors exhibit striking similarity to those observed in many natural systems, including biological neural networks. Experiment and numerical simulations have provided proof of principle that ASNs are complex systems whose emergent behaviors may be used in implementations of neuromorphic computing paradigms such as reservoir computing. Furthermore, they demonstrate the utility of ASNs as a uniquely scalable physical platform useful for exploring complexity, neuroscience, and engineering.

In situ hard x-ray photoemission spectroscopy (HX-PES) and soft x-ray absorption spectroscopy (SX-XAS) have been employed to investigate a local redox reaction at the carbon/Gd-doped CeO2 (GDC) thin film heterointerface under applied dc bias. In HX-PES, Ce3d and O1s core levels show a parallel chemical shift as large as 3.2 eV, corresponding to the redox window where ionic conductivity is predominant. The window width is equal to the energy gap between donor and acceptor levels of the GDC electrolyte. The Ce M-edge SX-XAS spectra also show a considerable increase of Ce3+ satellite peak intensity, corresponding to electrochemical reduction by oxide ion migration. In addition to the reversible redox reaction, two distinct phenomena by the electrochemical transport of oxide ions are observed as an irreversible reduction of the entire oxide film by O2 evolution from the GDC film to the gas phase, as well as a vigorous precipitation of oxygen gas at the bottom electrode to lift off the GDC film. These in situ spectroscopic observations describe well the electrochemical polarization behavior of a metal/GDC/metal capacitor-like two-electrode cell at room temperature.

MANA scientists discover that inorganic synapses mimic the human brain., MANA E-BULLETIN. 1 (2011)