Bitwise Neural Network Acceleration Using Silicon Photonics
Bitwise Neural Network Acceleration Using Silicon Photonics
复制标题
DOI：
10.1145/3453688.3461515
复制
发表时间：
2021-06
期刊：
影响因子：
--
通讯作者：
Kyle Shiflett;Avinash Karanth;A. Louri;Razvan C. Bunescu
中科院分区：
文献类型：
--
作者：
Kyle Shiflett;Avinash Karanth;A. Louri;Razvan C. Bunescu
硬件加速器为一些要求苛刻的深度神经网络（DNN）应用提供了显著的加速和提高能源效率。dnn有几个隐藏层，在网络权重和输入特征之间执行并发矩阵向量乘法（mvm）。由于mvm对深度神经网络的性能至关重要，以往的研究从架构和算法两个层面对mvm的性能和能效进行了优化。在本文中，我们提出利用新兴的硅光子技术来提高并行性、速度和整体效率，以提供神经网络的实时推理和快速训练。我们使用微环谐振器（MRRs）和马赫-曾德干涉仪（MZIs）设计了两种版本（全光学和部分光学）的混合矩阵乘法用于dnn。我们的研究结果表明，我们的部分光学设计在能源效率和延迟方面都具有最佳性能，保守估计的能量延迟产品（EDP）降低了33.1%，激进估计的EDP降低了76.4%。
Hardware accelerators provide significant speedup and improve energy efficiency for several demanding deep neural network (DNN) applications. DNNs have several hidden layers that perform concurrent matrix-vector multiplications (MVMs) between the network weights and input features. As MVMs are critical to the performance of DNNs, previous research has optimized the performance and energy efficiency of MVMs at both the architecture and algorithm levels. In this paper, we propose to use emerging silicon photonics technology to improve parallelism, speed and overall efficiency with the goal of providing real-time inference and fast training of neural nets. We use microring resonators (MRRs) and Mach-Zehnder interferometers (MZIs) to design two versions (all-optical and partial-optical) of hybrid matrix multiplications for DNNs. Our results indicate that our partial optical design gave the best performance in both energy efficiency and latency, with a reduction of 33.1% for energy-delay product (EDP) with conservative estimates and a 76.4% reduction for EDP with aggressive estimates.