• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Studia Universitatis Babes-Bolyai. Series Physica >CELLULAR NEURAL NETWORK COMPUTERS AND THEIRAPPLICATIONS IN PHYSICS
【24h】

CELLULAR NEURAL NETWORK COMPUTERS AND THEIRAPPLICATIONS IN PHYSICS

机译：细胞神经网络计算机及其在物理中的应用

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The computational paradigm represented by Cellular Neural/Nonlinear Networks(CNN) and the CNN Universal Machine (CNN-UM) as a Cellular Wave Computer, gives newperspectives for computational physics. Many numerical problems and simulations can be elegantlyaddressed on this fully parallelized and analogic architecture: solving partial differential equationsand implementing cellular automata models are just sonic basic examples. We also study thepossibility of Performing stochastic simulations on this chip. First a realistic random numbergenerator is implemented on the CNN-UM, then as an example the site-percolation problem and thetwo-dimensional Ising model are studied by Monte Carlo type simulations. The results obtained onan experimental version of the CNN-UM with 128×128 cells (ACEI6K) are in good agreementwith the results obtained on digital computers. Computational time measurements suggest that thedeveloping trend of the CNN-UM chips - increasing the lattice size and the number of local logicmemories - will assure an important advantage for the CNN-UM in the near future.
机译：以细胞神经/非线性网络（CNN）和CNN通用机（CNN-UM）为细胞波计算机所代表的计算范例，为计算物理学提供了新的视角。在这种完全并行化和类似的体系结构上，许多数值问题和模拟都可以很好地解决：求解偏微分方程和实现元胞自动机模型只是声音的基本示例。我们还研究了在该芯片上执行随机仿真的可能性。首先在CNN-UM上实现了一个现实的随机数生成器，然后以蒙特卡洛（Monte Carlo）型模拟为例研究了站点渗流问题和二维Ising模型。在具有128×128个单元的CNN-UM实验版本（ACEI6K）上获得的结果与在数字计算机上获得的结果非常吻合。计算时间测量表明，CNN-UM芯片的发展趋势-增加晶格大小和局部逻辑存储器的数量-将确保CNN-UM在不久的将来具有重要的优势。

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