【24h】

Optical matrix processing: a review

机译:光学矩阵处理:审查

获取原文

摘要

Matrix multiplication is well suited to optical techniques, since the size of the matrix has no direct influence on the speed with which the multiplication takes place. Although iterative techniques have been developed to solve systems of linear equations, to find the eigenvectors of matrices, and to perform other matrix manipulations, research has been hampered somewhat by a lack of fast two-dimensional spatial light modulators (SLMs). Advances in SLM technology are discussed, and different architectures for matrix processing are reported. Acoustic Bragg cells, liquid crystal devices, magnetooptic devices and silicon/PLZT crystal light modulators are briefly reviewed. Both Systolic and fully parallel optical vector-matrix multiplier configurations are described. Iterative optical processors and bimodal (analog-digital) optical computer accuracy are briefly considered.
机译:矩阵乘法非常适合光学技术,因为矩阵的大小对乘法的速度没有直接影响。尽管已经开发了迭代技术来解决线性方程的系统,但是找到矩阵的特征向量,并且执行其他矩阵操纵,在缺乏快速的二维空间光调制器(SLM)时,研究已经稍微受到困扰。讨论了SLM技术的进步,并报告了用于矩阵处理的不同架构。简要介绍声学布拉格电池,液晶装置,磁光装置和硅/ PLZT晶体光调制器。描述了Systolic和完全并行光学矢量矩阵乘法器配置。简要考虑迭代光学处理器和双峰(模拟数字)光学计算机精度。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号