General information
The race to realize all-optical intelligent machines, promising ultrafast and ultra-efficient computations for large-scale problems, continues. Optical computing has shown exceptional performance in areas from basic math to complex system optimization. In OPTICAL-BRAIN, we advance by integrating an active oscillatory medium into light scattering, adding temporal freedom and coupling it with nonlinear oscillators' dynamics. We will study nematic liquid crystals—soft matter with fluidity and long-range order—near bifurcation points described by the Ginzburg-Landau equation.
OPTICAL-BRAIN aims to develop an optical computing framework, leveraging continuous interaction dynamics between light and oscillatory matter, as an alternative to digital algorithms. This will enable: 1. Programmable light-induced soft oscillatory networks. 2. Neural networks operating solely in the optical domain. 3. Breakthroughs in scalability, power efficiency, and speed.
OPTICAL-BRAIN will combine soft matter physics, light-matter interaction, optics, neural network algorithms, and signal processing to introduce new optical methods and devices for analog computing, applicable to neural networks and combinatorial optimizations.
