# Compact Optical Reservoir Computing via Luminescence Dynamics in Rare‐Earth Ions‐Doped Nanocrystals

## TL;DR

This paper introduces a compact optical computing system using rare-earth ion-doped nanocrystals to enable efficient, real-time neuromorphic computing.

## Contribution

The novelty lies in using rare-earth ions' luminescence dynamics for nonlinear mapping and memory in a compact, optical reservoir computing system.

## Key findings

- The system achieved 90.7% accuracy in MNIST digit classification.
- Low-error chaotic time-series prediction was demonstrated with NRMSE < 0.1.
- The platform reduces system footprint and complexity compared to traditional optical computing methods.

## Abstract

Optical neuromorphic computing offers a promising route to high‐speed, energy‐efficient information processing. However, photonic neurons, as the critical components for enhancing computational expressivity, still face significant bottlenecks in nonlinear mapping and memory capacity. Here, a functionally compact optical reservoir computing system based on rare‐earth ions‐doped nanocrystals is demonstrated, leveraging their intrinsic nonlinear luminescence dynamics and multi‐timescale memory. Unlike traditional schemes that require bulky optical delays or intricate resonant structures, the platform exploits the material's inherent properties: nonlinear cross‐relaxation processes enable nonlinear mapping while long‐lived metastable energy levels provide fading memory. As a proof of concept, 90.7% accuracy is achieved in MNIST digit classification and low‐error chaotic time‐series prediction (NRMSE < 0.1) using the rare‐earth ions‐based system. This work significantly reduces system footprint and complexity, offering a scalable, fully optical solution for edge computing and real‐time neuromorphic applications.

A functionally compact optical reservoir computing platform is demonstrated using rare‐earth ion‐doped nanocrystals, whose intrinsic nonlinear luminescence dynamics and multi‐timescale metastable states provide nonlinear mapping and fading memory without bulky delays or resonant structures. This fully optical approach reduces system complexity and footprint, enabling scalable, energy‐efficient neuromorphic computing for real‐time edge applications.