卷积(计算机科学)
计算机科学
人工神经网络
光子学
硅光子学
最上等的
接头(建筑物)
人工智能
电子工程
计算科学
光学
光电子学
材料科学
工程类
物理
方位角
建筑工程
作者
Nicola Peserico,Shurui Li,Hangbo Yang,Hamed Dalir,Puneet Gupta,Volker J. Sorger
摘要
Convolution Neural Networks are one of the pillars of the Machine Learning revolution over the last years. However, convolution operation is associated with a high computational cost. Here, we present a compact and high-speed solution to perform convolution leveraging optics on a chip, with the potential to reach over 350 TOPS/W.
科研通智能强力驱动
Strongly Powered by AbleSci AI