Programmable chalcogenide-based all-optical deep neural networks.

We demonstrate a passive all-chalcogenide all-optical perceptron scheme. The network's nonlinear activation function (NLAF) relies on the nonlinear response of GeSbTe to femtosecond laser pulses. We measured the sub-picosecond time-resolved optical constants of GeSbTe at a wavelength of 1500 nm and used them to design a high-speed GeSbTe-tuned microring resonator all-optical NLAF. The NLAF had a sigmoidal response when subjected to different laser fluence excitation and had a dynamic range of -9.7 dB. The perceptron's waveguide material was AlN because it allowed efficient heat dissipation during laser switching. A two-temperature analysis revealed that the operating speed of the NLAF is ns. The percepton's nonvolatile weights were set using low-loss SbS-tuned Mach Zehnder interferometers (MZIs). A three-layer deep neural network model was used to test the feasibility of the network scheme and a maximum training accuracy of 94.5% was obtained. We conclude that combining SbS-programmed MZI weights with the nonlinear response of GeSbTe to femtosecond pulses is sufficient to perform energy-efficient all-optical neural classifications at rates greater than 1 GHz.