Ultra-Compact Inverse-Designed Integrated Photonic Matrix Compute Core
Open Access
Abstract
Leveraging our developed Global–Local Integrated Topology inverse design algorithm, we designed an efficient, compact, and symmetrical power splitter on a silicon-on-insulator platform. This device achieves a low insertion loss of 0.18 dB and a power imbalance of 80% accuracy with >0.9 probability under simulated random fabrication errors.Keywords
Funding Information
- National Key Research and Development Program of China (2021YFC2201902)
- National Natural Science Foundation of China (61975149)
This publication has 27 references indexed in Scilit:
- Ultra-compact multi-task processor based on in-memory optical computingLight: Science & Applications, 2025
- Optical neural networks: progress and challengesLight: Science & Applications, 2024
- Compact optical convolution processing unit based on multimode interferenceNature Communications, 2023
- Photonic machine learning with on-chip diffractive opticsNature Communications, 2023
- Space-efficient optical computing with an integrated chip diffractive neural networkNature Communications, 2022
- Inverse design in nanophotonicsNature Photonics, 2018
- Inverse Design and Demonstration of a Compact on-Chip Narrowband Three-Channel Wavelength DemultiplexerACS Photonics, 2017
- Deep learning with coherent nanophotonic circuitsNature Photonics, 2017
- An integrated-nanophotonics polarization beamsplitter with 2.4 × 2.4 μm2 footprintNature Photonics, 2015
- Inverse design and demonstration of a compact and broadband on-chip wavelength demultiplexerNature Photonics, 2015