Proposal for nanoscale cascaded plasmonic majority gates for non-Boolean computation

TL;DR

This paper introduces a nanoscale plasmonic majority gate that uses wave phase instead of intensity for computation, enabling non-Boolean logic and potential applications in real-time signal processing.

Contribution

It proposes a novel phase-based plasmonic majority gate with a cascadable design and a referencing scheme, advancing beyond traditional intensity-based plasmonic logic devices.

Findings

Demonstrates a 3-input plasmonic majority gate with 0.636 μm² area

Uses phase of surface-plasmon-polariton waves as the computational variable

Enables non-Boolean computation suitable for pattern recognition

Abstract

Surface-plasmon-polariton waves propagating at the interface between a metal and a dielectric, hold the key to future high-bandwidth, dense on-chip integrated logic circuits overcoming the diffraction limitation of photonics. While recent advances in plasmonic logic have witnessed the demonstration of basic and universal logic gates, these CMOS oriented digital logic gates cannot fully utilize the expressive power of this novel technology. Here, we aim at unraveling the true potential of plasmonics by exploiting an enhanced native functionality - the majority voter. Contrary to the state-of-the-art plasmonic logic devices, we use the phase of the wave instead of the intensity as the state or computational variable. We propose and demonstrate, via numerical simulations, a comprehensive scheme for building a nanoscale cascadable plasmonic majority logic gate along with a novel referencing scheme that can directly translate the information encoded in the amplitude and phase of the wave into electric field intensity at the output. Our MIM-based 3-input majority gate displays a highly improved overall area of only 0.636 {\mu}m$^2$ for a single-stage compared with previous works on plasmonic logic. The proposed device demonstrates non-Boolean computational capability and can find direct utility in highly parallel real-time signal processing applications like pattern recognition.