APPLICATIONS OF TECHNOLOGY:
- Energy efficient all optical control (i.e. optical transistors)
- All-optical neural networks
- Analog computing based on light
- Real-time image processing
BENEFITS:
- Energy efficient
- Highly parallel
- Compatible with incoherent light such as LEDs and ambient light.
BACKGROUND:
Optical computing has been sought after for many decades because it allows for massively parallel information processing. Optical neural networks use linear optics to implement fundamental mathematical operations which are the building blocks of inference tasks. However, the widespread adoption of all optical neural networks is hindered by the lack of energy efficient optical nonlinearities that are required for deep neural networks. Existing technology requires high power coherent lasers to realize optical nonlinearity, which is incompatible with energy-efficient and large-scale optical computation.
TECHNOLOGY OVERVIEW:
Berkeley Lab scientists have developed a new nonlinear optical microdevice array that enables low energy and massively parallel nonlinear optical control of incoherent light. The new device architecture integrates a liquid-crystal (LC) electro-optical modulator and silicon pn-junctions at a single pixel level, which is fully compatible with existing industrial-level manufacturing. The device capitalizes on the electro-optical properties of LC and the integrated photodiodes to modulate incident light, thereby offering control over phase, polarization, and intensity of a single beam.
DEVELOPMENT STAGE:
Engineering/pilot-scale, prototypical system validation in relevant environment.
PRINCIPAL INVESTIGATORS:
Feng Wang
Qixin Feng
Can Uzundal
IP Status:
Patent pending
OPPORTUNITIES:
Available for licensing or collaborative research