Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 1808.05135

Neuromorphic computing in Ginzburg-Landau polariton lattice systems

Andrzej Opala·Sanjib Ghosh·Timothy C. H. Liew·Michał Matuszewski
View original source ↗Read source document ↗Export metadata ↓

Abstract

The availability of large amounts of data and the necessity to process it efficiently have led to rapid development of machine learning techniques. To name a few examples, artificial neural network architectures are commonly used for financial forecasting, speech and image recognition, robotics, medicine, and even research. Direct hardware for neural networks is highly sought for overcoming the von Neumann bottleneck of software implementations. Reservoir computing (RC) is a recent and increasingly popular bio-inspired computing scheme which holds promise for an efficient temporal information processing. We demonstrate the applicability and performance of reservoir computing in a general complex Ginzburg-Landau lattice model, which adequately describes dynamics of a wide class of systems, including coherent photonic devices. In particular, we propose that the concept can be readily applied in exciton-polariton lattices, which are characterized by unprecedented photonic nonlinearity, opening the way to signal processing at rates of the order of 1 Tbit (1/s).

Explore related subjects

cond-mat.mes-hallcond-mat.quant-gas

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Andrzej Opala, Sanjib Ghosh, Timothy C. H. Liew, Michał Matuszewski. 2019-05-19. Neuromorphic computing in Ginzburg-Landau polariton lattice systems. https://doi.org/10.1103/physrevapplied.11.064029

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Substrate-driven topological engineering in plasmonic Su-Schrieffer-Heeger chains

We demonstrate the possibility of engineering the topological band structure of a plasmonic Su-Schrieffer-Heeger (SSH) chain through the interaction with its electromagnetic environment. We find that the long-range interaction of the in-plane modes of the SSH chain with the surface plasmon polaritons of a planar substrate introduces a band hybridization connected to a change of the Zak phase. On the other hand, the short-range interaction with the substrate introduces a band touching, again with a change in the Zak phase. Surprisingly, this second mechanism enables the emergence of topologically protected edge modes for parameters which correspond to the topologically trivial phase for an isolated plasmonic SSH chain. We study these mechanisms by changing the chain-substrate distance and the dimerization parameter. Finally, we discuss the robustness against disorder and, as one example, the impact of the observed effects on the near-field radiative heat transfer and the local density of states along the chain. Our findings pave the way to the engineering of edge modes in plasmonic topological configurations via the coupling to a plasmonic environment.

cond-mat.mes-hall↗
arXiv2026

Hysteresis-Driven Radiative Mpemba Effect in Phase-Change Nanostructures

The Mpemba effect states that initially hotter systems cool faster than colder ones. While known in convective, conductive, and quantum systems, its radiative analogue is unexplored. Here, this anomaly is realized via phase-change hysteresis of a VO$_2$ nanoparticle near a SiC substrate. After analytically deriving an onset condition, the phase space is mapped. Crucially, latent heat acts as a thermal buffer enabling both ordinary and inverse effects. Near-field coupling governs the relaxation time and enables a passive effect where memory is stored externally via substrate reflection.

cond-mat.mes-hall↗
arXiv2026

Fast universal parametric spin control in an acoustically modulated quantum dot

Quantum communication, distributed computing, and hybrid architectures rely on nodes enabling coherent control of qubits and coupling to propagating quantum modes. While semiconductor quantum-dot (QD) spins couple to microwave and optical photons, weak interaction with mechanical waves has limited the integration of single-QD spin qubits into on-chip, acoustically coupled hybrid systems. The existing theory of acoustic QD spin control suffers from a limited range of rotation-axis angles, enforcing complex realizations of gate primitives and long gate times, leaving little margin against decoherence from trion decay and quasi-static nuclear-spin noise. We propose parametric control that overcomes these problems. We use far-detuned optical coupling to a trion state to dress and thus mix spin states, combined with acoustic modulation of the optical transition energy. Instead of relying on direct acoustic resonance with the spin splitting that leads to significant bottlenecks, we induce spin rotations parametrically via resonance with the dressed-spin splitting. We thus develop a spin analog of the ``swing-up'' charge-state excitation. Our scheme provides fast universal qubit control with nearly arbitrary rotation axes. Our ${\sim}$155 ps Pauli-$X$ gate duration is ${\sim}290\times$ faster than in the previous acousto-optical formulation and ${\sim}14\times$ faster than optical Faraday-geometry spin rotation. The parametric scheme naturally enables higher-harmonic processes. Numerical simulations for ${\sim}44$ GHz acoustic driving show average gate fidelity $\ge99.9\%$ even for uncooled nuclear-spin environments of GaAs and InAs QDs for trion lifetime $\gtrsim1.25$ ns. These metrics suggest practically usable control and may introduce a spin-phonon interface with high interaction rates, versatility, and multi-phonon processes, essential for future acoustically coupled hybrid architectures.

cond-mat.mes-hall↗