OPTICAL NEURAL NETWORK ACCELERATOR

Organization Name

intel corporation

Inventor(s)

Songtao Liu of Santa Clara CA (US)

Haisheng Rong of Pleasanton CA (US)

Mozhgan Mansuri of Portland OR (US)

Ram Krishnamurthy of Portland OR (US)

OPTICAL NEURAL NETWORK ACCELERATOR - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240242067 titled 'OPTICAL NEURAL NETWORK ACCELERATOR

The patent application describes technology that performs matrix-matrix multiplication operations in an optical neural network (ONN) using a first plurality of panels to generate output optical signals based on input optical signals and weights.

The technology includes an input panel, a weight panel, and a photodetector panel within the first plurality of panels.
The input panel generates input optical signals for the matrix-matrix multiplication operation.
The weight panel represents the weights of the first layer of the ONN.
The photodetector panel generates output photodetector signals based on the output optical signals produced by the ONN.

Potential Applications: - Optical computing - Neural network processing - Image recognition systems

Problems Solved: - Efficient matrix-matrix multiplication in optical neural networks - Improved performance of neural network operations

Benefits: - Faster processing speeds - Reduced energy consumption - Enhanced accuracy in neural network computations

Commercial Applications: Title: Optical Neural Network Technology for Advanced Computing Systems This technology can be applied in industries such as: - Data centers - Autonomous vehicles - Healthcare imaging systems

Prior Art: Researchers can explore prior studies on optical neural networks and matrix-matrix multiplication in optical systems to understand the existing knowledge in this field.

Frequently Updated Research: Stay updated on advancements in optical computing, neural network algorithms, and photonic technologies to enhance the performance of ONNs.

Questions about Optical Neural Network Technology: 1. How does this technology improve the efficiency of matrix-matrix multiplication in neural networks? 2. What are the potential limitations of implementing optical neural networks in practical applications?

Original Abstract Submitted

systems, apparatuses and methods include technology that executes, with a first plurality of panels, a first matrix-matrix multiplication operation of a first layer of an optical neural network (onn) to generate output optical signals based on input optical signals that pass through an optical path of the onn, and weights of the first layer of the onn. the first plurality of panels includes an input panel, a weight panel and a photodetector panel. the executing includes generating, with the input panel, the input optical signals, where the input optical signals represent an input to the first matrix-matrix multiplication operation of the first layer of the onn, representing, with the weight panel, the weights of the first layer of the onn, and generating, with the photodetector panel, output photodetector signals based on the output optical signals that are generated based on the input optical signals and the weights.