Programmable chalcogenide-based all-optical deep neural networks
Teo Ting Yu,
Ma Xiaoxuan,
Pastor Ernest,
Wang Hao,
George Jonathan K.,
Yang Joel K. W.,
Wall Simon,
Miscuglio Mario,
Simpson Robert E.,
Sorger Volker J.
Affiliations
Teo Ting Yu
Singapore University of Technology and Design, 8 Somapah Road, Singapore487372, Singapore
Ma Xiaoxuan
Deptartment of Electrical and Computer Engineering, George Washington University, Washington, DC, USA
Pastor Ernest
ICFO - Institut de Ciencies Fotoniques, The Barcelona Institute of Science and Technology, Av. Carl Friedrich Gauss 3, Castelldefels08860, Barcelona, Spain
Wang Hao
Singapore University of Technology and Design, 8 Somapah Road, Singapore487372, Singapore
George Jonathan K.
Deptartment of Electrical and Computer Engineering, George Washington University, Washington, DC, USA
Yang Joel K. W.
Singapore University of Technology and Design, 8 Somapah Road, Singapore487372, Singapore
Wall Simon
ICFO - Institut de Ciencies Fotoniques, The Barcelona Institute of Science and Technology, Av. Carl Friedrich Gauss 3, Castelldefels08860, Barcelona, Spain
Miscuglio Mario
Deptartment of Electrical and Computer Engineering, George Washington University, Washington, DC, USA
Simpson Robert E.
Singapore University of Technology and Design, 8 Somapah Road, Singapore487372, Singapore
Sorger Volker J.
Deptartment of Electrical and Computer Engineering, George Washington University, Washington, DC, USA
We demonstrate a passive all-chalcogenide all-optical perceptron scheme. The network’s nonlinear activation function (NLAF) relies on the nonlinear response of Ge2Sb2Te5 to femtosecond laser pulses. We measured the sub-picosecond time-resolved optical constants of Ge2Sb2Te5 at a wavelength of 1500 nm and used them to design a high-speed Ge2Sb2Te5-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 ≤1 $\le 1$ ns. The percepton’s nonvolatile weights were set using low-loss Sb2S3-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 Sb2S3-programmed MZI weights with the nonlinear response of Ge2Sb2Te5 to femtosecond pulses is sufficient to perform energy-efficient all-optical neural classifications at rates greater than 1 GHz.
