摘要
Optical approaches have made great strides towards the goal of high-speed, energy-efficient computing necessary for modern deep learning and AI applications. Read-in and read-out of data, however, limit the overall performance of existing approaches. This study introduces a multilayer optoelectronic computing framework that alternates between optical and optoelectronic layers to implement matrix-vector multiplications and rectified linear functions, respectively. Our framework is designed for real-time, parallelized operations, leveraging 2D arrays of LEDs and photodetectors connected via independent analog electronics. We experimentally demonstrate this approach using a system with a three-layer network with two hidden layers and operate it to recognize images from the MNIST database with a recognition accuracy of 92% and classify classes from a nonlinear spiral data with 86% accuracy. By implementing multiple layers of a deep neural network simultaneously, our approach significantly reduces the number of read-ins and read-outs required and paves the way for scalable optical accelerators requiring ultra low energy.
学科与关键词
- An optical neural network using less than 1 photon per multiplicationWang, Tianyu*; Ma, Shi-Yuan; Wright, Logan G.; Onodera, Tatsuhiro; Richard, Brian C.; McMahon, Peter L.*Nature Communications, 2022, 13(1): 123.被引 122
- Light in AI: Toward Efficient Neurocomputing With Optical Neural Networks-A TutorialGu, Jiaqi*; Feng, Chenghao; Zhu, Hanqing; Chen, Ray T.; Pan, David Z.IEEE Transactions on Circuits and Systems II: Express Briefs , 2022, 69(6): 2581-2585.被引 11
- Freely scalable and reconfigurable optical hardware for deep learningBernstein, Liane*; Sludds, Alexander*; Hamerly, Ryan; Sze, Vivienne; Emer, Joel; Englund, Dirk*Scientific Reports, 2021, 11(1): 3144.被引 43
