announcement | Wed, Aug 19th, 2020 facebook twitter linkedin Email IEEE Journal on Exploratory Solid-State Computational Devices and Circuits Special Issue on Coupled Oscillators for Non- von Neumann Computation When oscillators are loosely coupled to each other, energy transfer between the individual oscillators causes their frequencies to synchronize. The same principle can be found in real life; for instance, metronomes placed on a floating wooden board, pendulums connected via springs, and internal organs following a circadian rhythm. Depending on the strength and time lag of the coupling medium, the phases of the oscillators settle in a way that minimizes the contentions among the oscillating signals. Recent works have shown that the coupled oscillator’s natural ability to evolve to the ground state can be exploited to solve computationally intractable problems, such as graph coloring, max cut, factorization, neural networks, associative memories, and pattern recognition. Here, the problems are first mapped to a coupled oscillator network by configuring the coupling weights, and the phase information is read out once the ground state is found. While resolving to the ground state, the network may get stuck in a local minima state, which can be avoided by a concept called annealing where random noise is added during the early exploration phase to help the oscillators break out of a local minimal state. Coupled oscillator networks vary in their device implementation as well as in their connectivity. For the devices, experimental demonstrations include CMOS oscillators, emerging device-based, such as ferroelectric, spintronic, phase change oscillators, optical oscillators, and quantum devices at cryogenic temperatures. In some cases, oscillators were discrete devices assembled on a board, in other cases, they were monolithically integrated on a chip. In terms of connectivity, fully-connected, nearest-neighbor, hybrid networks (e.g. Chimera), and common node coupling architectures have been demonstrated. Against this backdrop, the IEEE Journal on Exploratory Computational Devices and Circuits (JXCDC) is pleased to announce the next special issue focusing all aspects of coupled oscillator based system specifically targeted for non-von Neumann computing applications. Topics of Interest Emerging device (e.g. optical, NEMS, ferroelectric, spintronic, phase change) based coupled oscillator systems CMOS based coupled oscillator systems Variability and reliability effects in coupled oscillator systems Probabilistic behavior and operation under noise Security properties of coupled oscillator systems Weight programming and phase readout techniques Annealing techniques for coupled oscillator systems Network connectivity and architecture considerations Testing, parameter turning, and measurements aspectS Oscillator Neural Networks (ONNs) Associative memories based on oscillators Techniques for mapping large problems onto coupled oscillator systems Graph embedding algorithms for locally connected coupled oscillator systems NP-hard and NP-complete problem case studies Comparison with quantum computers and software based approaches (e.g. simulated annealing) Literature review and historical trends on coupled oscillator systems Important Dates Open for Submission: July 10, 2020 Submission Deadline: September 30, 2020 First Notification: Ocotber 21, 2020 Revision Submission: November 15, 2020 Final Decision: December 15, 2020 Publication Online: January 1, 2021 View the full Call for Papers Related Content CFP: Special Issue on Coupled Oscillators for Non- von Neumann Computation
