Silicon Chip Distributes Optical Signals for Potential Use in Neural Networks

A silicon chip that distributes optical signals precisely across a miniature grid could lead to a novel design for neural networks.

A research team from the National Institute of Standards and Technology (NIST) proposes to use light, rather than conventional electronics, as a signaling medium for artificial neural networks. The use of light would eliminate interference due to electrical charge, and the light signals would travel faster and farther.

NIST's grid-on-a-chip distributes light signals precisely, showcasing a potential new design for neural networks. The three-dimensional structure enables complex routing schemes, which are necessary to mimic the brain. Light could travel farther and faster than electrical signals. Courtesy of Chiles/NIST.

Using two vertically integrated planes of silicon nitride waveguides, the NIST team designed, fabricated, and characterized integrated photonic routing manifolds. By vertically stacking two layers of waveguides, the team created a 3D design that could enable complex routing schemes. Moreover, the design can be extended to incorporate additional waveguiding layers for even more complex networks. The 3D grid has 10 inputs or “upstream” neurons, each connecting to 10 outputs or “downstream” neurons, for a total of 100 receivers.

Fabricated on a silicon wafer, the waveguides are made of silicon nitride and are each 800 nm wide and 400 nm thick. Researchers created software to automatically generate signal routing, with adjustable levels of connectivity between the neurons.

Laser light was directed into the chip through an optical fiber. The goal was to route each input to every output group, following a selected distribution pattern for light intensity. Researchers demonstrated two schemes for controlling output intensity: uniform, where each output receives the same power; and a "bell curve" distribution, where middle neurons receive the most power, while peripheral neurons receive less.

Researchers analyzed the output signals via top-view camera imaging. All signals were focused through a microscope lens onto a semiconductor sensor and processed into image frames. This technique allowed the rapid acquisition of hundreds of precise transmission measurements.

The team demonstrated signals with uniform and Gaussian power distribution patterns with mean power output errors (averaged over 10 sets of 10 inputs) of 0.7 and 0.9 dB, respectively, establishing its 3D grid as a viable architecture for precision light distribution on-chip.

Researchers also assessed the performance of the passive photonic elements that constitute the system using self-referenced test structures, including high-dynamic-range beam taps, waveguide cutback structures, and waveguide crossing arrays.

“We’ve really done two things here,” researcher Jeff Chiles said. “We’ve begun to use the third dimension to enable more optical connectivity, and we’ve developed a new measurement technique to rapidly characterize many devices in a photonic system. Both advances are crucial as we begin to scale up to massive optoelectronic neural systems.”

The team believes that these routing and measurement techniques could offer new opportunities for complex integrated photonic systems in computing, telecommunications, and other applications.

“Light’s advantages could improve the performance of neural nets for scientific data analysis such as searches for Earth-like planets and quantum information science, and accelerate the development of highly intuitive control systems for autonomous vehicles,” Chiles said.

The research was published in APL Photonics (doi: 10.1063/1.5039641).