Berkeley Lab Scientists Invent Novel Microdevice Array for Energy Efficient Optical Computing

Scientists at Lawrence Berkeley National Laboratory have developed a breakthrough technology that could greatly develop the field of optical computing and devices utilizing all- optical control. The invention, developed by Berkeley Lab scientists Feng Wang, Qixin Feng, and Can Uzundal, is a microdevice array capable of efficiently performing complex and simultaneous analog computations using photons. 

This breakthrough not only opens new avenues in optical computing but also holds immense promise for a spectrum of applications demanding high-speed data processing and computational efficiency. This device also offers improvements for devices using energy efficient all-optical control, such as optical transistors, switches, and modulators.

Optical computing has long been pursued for its potential to enable multiple computations simultaneously. Optical neural networks, in particular, leverage linear optics to execute fundamental mathematical operations crucial for inference tasks such as pattern recognition, classification, and other data analysis techniques. However, existing technology relies on high-power coherent lasers to achieve optical nonlinearity needed for optical computing, making it highly energy intensive and rendering it incompatible with large-scale optical computation and real-life scenarios.

This cutting-edge technology developed by Berkeley Lab addresses these challenges head-on. The new microdevice array showcases a sophisticated architecture, integrating a liquid-crystal (LC) electro-optical modulator and silicon pn-junctions at the single pixel level. At its core, the device harnesses the electro-optical properties of LCs and the integrated photodiodes to effectively control incident light. 

By precisely manipulating phase, polarization, or intensity of individual light beams, this advancement marks a significant leap forward in optical computing capabilities. The design achieved highly parallel optical nonlinearity with millions of pixels (translating to orders-of-magnitude speed-up), high energy efficiency at sub-pico Joule level, and is compatible with incoherent LED light. Remarkably, the scientists employed the microdevice array to exhibit an optical analogue of the Rectified Linear Unit (ReLU) function, which is widely utilized in deep neural networks. These achievements suggest that the invention holds promise for developing ambient light-based deep computing hardware, with potential applications in various fields, including the automotive industry, virtual/augmented reality (VR/AR), artificial intelligence/machine learning, and  edge computing.

This innovation aligns seamlessly with existing industrial-level manufacturing processes, paving the way for widespread accessibility and adoption. 

The technology is now available for licensing; contact [email protected]. For more information on optical computing innovations from Berkeley Lab, please visit the IPO website.

The research for this technology was funded by the U.S. Department of Energy.

###

Lawrence Berkeley National Laboratory (Berkeley Lab) is committed to delivering solutions for humankind through research in clean energy, a healthy planet, and discovery science. Founded in 1931 on the belief that the biggest problems are best addressed by teams, Berkeley Lab and its scientists have been recognized with 16 Nobel Prizes. Researchers from around the world rely on the Lab’s world-class scientific facilities for their own pioneering research. Berkeley Lab is a multiprogram national laboratory managed by the University of California for the U.S. Department of Energy’s Office of Science.

DOE’s Office of Science is the single largest supporter of basic research in the physical sciences in the United States, and is working to address some of the most pressing challenges of our time. For more information, please visit energy.gov/science.