{"title":"A Guided-Wave Acoustooptic Matrix Algebra Processor Module*","authors":"A. Kar-Roy, C. Tsai","doi":"10.1364/optcomp.1991.tud3","DOIUrl":null,"url":null,"abstract":"Various architectures for implementing matrix algebra processors (MAPs) have been proposed and developed in bulk-wave optical systems(1−7) as well as in integrated optic (IO) devices(8−10) Bulk-wave optical systems have the advantage of an added dimension in implementing 3-D architectures over planar IO devices in which only 2-D architectures are possible. But the IO devices have the potential advantages in terms of drive power requirement, size, robustness, stability and planar technology for mass production. In this paper, we report on a guided-wave acoustooptic (AO) analog MAP module that is capable of implementing high-speed matrix-vector and matrix-matrix multiplications. The architecture, the working principle, and a preliminary MAP constructed on a 1.0 X 10.0 X 28.0 mm3 Y-cut LiNbO3 substrate to demonstrate the multiplication of a 4 X 4 matrix with a 4-element vector, are presented.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"80 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1992-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Optical Computing","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1364/optcomp.1991.tud3","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Various architectures for implementing matrix algebra processors (MAPs) have been proposed and developed in bulk-wave optical systems(1−7) as well as in integrated optic (IO) devices(8−10) Bulk-wave optical systems have the advantage of an added dimension in implementing 3-D architectures over planar IO devices in which only 2-D architectures are possible. But the IO devices have the potential advantages in terms of drive power requirement, size, robustness, stability and planar technology for mass production. In this paper, we report on a guided-wave acoustooptic (AO) analog MAP module that is capable of implementing high-speed matrix-vector and matrix-matrix multiplications. The architecture, the working principle, and a preliminary MAP constructed on a 1.0 X 10.0 X 28.0 mm3 Y-cut LiNbO3 substrate to demonstrate the multiplication of a 4 X 4 matrix with a 4-element vector, are presented.
在块波光学系统(1−7)以及集成光学(IO)器件(8−10)中,已经提出并开发了用于实现矩阵代数处理器(map)的各种体系结构。块波光学系统在实现三维体系结构方面具有比平面IO器件增加维度的优势,其中只有二维体系结构是可能的。但在驱动功率要求、尺寸、稳健性、稳定性和大规模生产的平面技术等方面,IO器件具有潜在的优势。在本文中,我们报告了一种能够实现高速矩阵-矢量和矩阵-矩阵乘法的导波声光(AO)模拟MAP模块。介绍了结构、工作原理和在1.0 X 10.0 X 28.0 mm3 y型切割LiNbO3衬底上构建的初步MAP,以演示4 X 4矩阵与4元素向量的乘法。