Details
| Originalsprache | Englisch |
|---|---|
| Seiten (von - bis) | 5080-5086 |
| Seitenumfang | 7 |
| Fachzeitschrift | ACS Omega |
| Jahrgang | 10 |
| Ausgabenummer | 5 |
| Frühes Online-Datum | 27 Jan. 2025 |
| Publikationsstatus | Veröffentlicht - 11 Feb. 2025 |
Abstract
The inverse design of photonic integrated circuits (PICs) presents distinctive computational challenges, including their large memory requirements. Advancements in the two-photon polymerization (2PP) fabrication process introduce additional complexity, necessitating the development of more flexible optimization algorithms to enable the creation of multimaterial 3D structures with unique properties. This paper presents a memory efficient reverse-mode automatic differentiation framework for finite-difference time-domain (FDTD) simulations that can handle complex constraints arising from novel fabrication methods. Our method is based on straight-through gradient estimation that enables nondifferentiable shape parametrizations. We demonstrate the effectiveness of our approach by creating increasingly complex structures to solve the coupling problems in PICs. The results highlight the potential of our method for future PIC design and practical applications.
ASJC Scopus Sachgebiete
- Chemie (insg.)
- Allgemeine Chemie
- Chemische Verfahrenstechnik (insg.)
- Allgemeine chemische Verfahrenstechnik
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
in: ACS Omega, Jahrgang 10, Nr. 5, 11.02.2025, S. 5080-5086.
Publikation: Beitrag in Fachzeitschrift › Artikel › Forschung › Peer-Review
}
TY - JOUR
T1 - Quantized Inverse Design for Photonic Integrated Circuits
AU - Schubert, Frederik
AU - Mahlau, Yannik
AU - Bethmann, Konrad
AU - Hartmann, Fabian
AU - Caspary, Reinhard
AU - Munderloh, Marco
AU - Ostermann, Jörn
AU - Rosenhahn, Bodo
N1 - Publisher Copyright: © 2025 The Authors. Published by American Chemical Society.
PY - 2025/2/11
Y1 - 2025/2/11
N2 - The inverse design of photonic integrated circuits (PICs) presents distinctive computational challenges, including their large memory requirements. Advancements in the two-photon polymerization (2PP) fabrication process introduce additional complexity, necessitating the development of more flexible optimization algorithms to enable the creation of multimaterial 3D structures with unique properties. This paper presents a memory efficient reverse-mode automatic differentiation framework for finite-difference time-domain (FDTD) simulations that can handle complex constraints arising from novel fabrication methods. Our method is based on straight-through gradient estimation that enables nondifferentiable shape parametrizations. We demonstrate the effectiveness of our approach by creating increasingly complex structures to solve the coupling problems in PICs. The results highlight the potential of our method for future PIC design and practical applications.
AB - The inverse design of photonic integrated circuits (PICs) presents distinctive computational challenges, including their large memory requirements. Advancements in the two-photon polymerization (2PP) fabrication process introduce additional complexity, necessitating the development of more flexible optimization algorithms to enable the creation of multimaterial 3D structures with unique properties. This paper presents a memory efficient reverse-mode automatic differentiation framework for finite-difference time-domain (FDTD) simulations that can handle complex constraints arising from novel fabrication methods. Our method is based on straight-through gradient estimation that enables nondifferentiable shape parametrizations. We demonstrate the effectiveness of our approach by creating increasingly complex structures to solve the coupling problems in PICs. The results highlight the potential of our method for future PIC design and practical applications.
KW - physics.optics
KW - physics.comp-ph
UR - http://www.scopus.com/inward/record.url?scp=85216201891&partnerID=8YFLogxK
U2 - 10.1021/acsomega.4c10958
DO - 10.1021/acsomega.4c10958
M3 - Article
VL - 10
SP - 5080
EP - 5086
JO - ACS Omega
JF - ACS Omega
IS - 5
ER -