Optical Computing
Abstract
We are exploring scalable all-optical solutions to minimize Ising Hamiltonians. These machines offer potential for solving NP-hard problems. Our interest is in (1) a metaphoric approach based on the nonlinear dynamics of coupled multi-core fiber lasers and (2) an algorithmic approach based on message passing and graphical inference.
Another topic of this program is an optical emulator for complex radiofrequency systems. Considering the difficulty to measure and to simulate RF properties of large scale structures, we can scale down the size of the object of interest and make measurement in the optical domain. The advantage is compactness of the system and a fast turn around in the production of the model via 3D printing. Radar cross section, antenna gain and interferences between antennas are the type of information it is possible to extract.
We gladly acknowledge the support from the Office of Naval Research for this project.
Publications:
- M. Babaeian et al., “Optical Versus Electronic Implementation of Probabilistic Graphical Inference and Experimental Device Demonstration Using Nonlinear Photonics“, IEEE Photonics Journal, vol. 10, Issue 5, (2018). Link
- M. Babaeian et al., “Nonlinear optical components for all-optical probabilistic graphical model”, Nature Communications, 9, 2128 (2018). Link
- P.-A. Blanche et al., “A 100,000 Scale Factor Radar Range“, Scientific Reports, 7, 17767, (2017). Link
- P.-A. Blanche et al., “All-Optical Graphical Models for Probabilistic Inference“, Invited paper at the IEEE Summer Topical Meeting on “Photonic Hardware Accelerators and Neuro-inspired Computing”, July 2016. Link.
- D. Nguyen et al., “An Optical Ising Machine Based on Multi-core Fiber Lasers“, Invited paper at the IEEE Summer Topical Meeting on “Photonic Hardware Accelerators and Neuro-inspired Computing”, July 2016. Link