Realization of optical logic gates using on-chip diffractive optical neural networks.

Area of Science:

• Photonics and Nanotechnology
• Computer Science

Background:

• Semiconductor electronic devices face performance limitations.
• Optical computing offers a potential solution for high-speed computation.
• Optical logic gates are essential for optical computation, enabling fast, low-heat, and crosstalk-free operations.

Purpose of the Study:

• To design a multi-functional optical logic gate.
• To demonstrate AND, NOT, and OR logic operations.
• To investigate broadband and wavelength-independent performance for parallel computation.

Main Methods:

• Design of an on-chip diffractive optical neural network.
• Implementation of logic operations at 1.55 µm wavelength.
• Testing wavelength-independent operation over a 60 nm bandwidth.

Main Results:

• A multi-functional optical logic gate capable of AND, NOT, and OR operations was designed.
• The device demonstrated wavelength-independent operation across seven wavelengths.
• The gate operates efficiently at 1.55 µm with a 60 nm bandwidth.

Conclusions:

• The developed optical logic gate is simple, highly integrable, low-loss, and energy-efficient.
• This broadband device offers a pathway for high-speed on-chip nanophotonic processors.
• It enables parallel computation through wavelength division multiplexing for future optical computing applications.