Summary
International Symposium on Nonlinear Theory and its Applications
2017
Session Number:C2L-D
Session:
Number:C2L-D-2
Optimizing Dynamical States for High-Speed Photonic Computing
Michael Wishon, Alexandre Locquet, Chien-Yuan Chang, Daeyoung Choi, David Citrin,
pp.685-687
Publication Date:2017/12/4
Online ISSN:2188-5079
Summary:
Photonic neuromorphic computing is the intersection of sub nanosecond fluctuations with the performance and energy efficiency of the human brain. These systems operate based on a nonlinear transformation of input streams, and by fully understanding the attractor dynamics of the transformation it is possible to optimize these dynamics for a computing task. By optimizing the dynamical state of our neural net, we are able to process over 70 million digits per second.