Figure 3
From: Freely scalable and reconfigurable optical hardware for deep learning
Possible implementation of digital optical neural network. (a)Â Digital inputs and weights are transmitted electronically to an array of light sources (red and blue, respectively, illustrating different paths). Single-mode light from a source is collimated by a spherical lens (Lens), then focused to a 1D spot array by a diffractive optical element (DOE). A 50:50 beamsplitter brings light from the inputs and weights into close proximity on a custom CMOS receiver. (b)Â Example circuit with 2 photodetectors (biased by voltage \(V_\text {bias}\)) per PE: 1 for activations; 1 for weights. Received bits (\(V_\text {out}\)) proceed to multiplier, then memory or next layer.
