首页> 外文期刊>Journal of Theoretical and Applied Information Technology >ABOUT THE FUNCTIONING OF THE SPECIAL-PURPOSE CALCULATING UNIT BASED ON THE LINEAR SYSTEM SOLUTION USING THE FIRST ORDER DELTA-TRANSFORMATIONS
【24h】

ABOUT THE FUNCTIONING OF THE SPECIAL-PURPOSE CALCULATING UNIT BASED ON THE LINEAR SYSTEM SOLUTION USING THE FIRST ORDER DELTA-TRANSFORMATIONS

机译:关于基于一阶DELTA变换的线性系统解决方案的特殊计算单元的功能

获取原文
获取外文期刊封面目录资料

摘要

This paper discusses the theoretical representations of special-purpose calculating unit functioning for the iteration solution of linear systems using the first order delta-transformations and variable quantum. It is considered the algorithm for iterative solution of linear systems based on the first order delta-transformations and variable quantum is adapted for implementation in a special-purpose calculating unit. A special feature of special-purpose calculating unit functioning based on this algorithm is the implementation of the introduction at the beginning of each cycle l of a new variable quantum value that is reflected in the current cycle when the residual values and the unknown variable are formed by shifting them to the left by 1 or 2 bits. Formation of unknown variables in the unit is carried out by adding or subtracting the signs of quantum of the first and second variables differences, taking at each iteration the values ? 1. This feature of variable normalization represents the possibility of organizing a computational process on the basis of an integer data representation. At the final step of the algorithm operation in the unit, it is possible to bring the values of the variables to the original real form, taking into account the weight of the minimum transformation quantum. With the orientation to FPGA, comparative estimates are obtained for the hardware and time resources of the developed algorithm and comprehensive comparative estimate of the effectiveness for special-purpose calculating unit functioning. In this paper for the developed algorithm of the unit functioning, it is shown that it is possible to reduce the execution of one iteration and the iterative process as a whole, the amount of hardware resources and generally improve the efficiency in comparison with the special-purpose calculating unit functioning based on the simple iteration method.
机译:本文讨论了使用一阶delta变换和可变量子的线性计算系统迭代解的专用计算单元的理论表示。它被认为是基于一阶delta变换的线性系统的迭代求解算法,并且可变量子适合在专用计算单元中实现。基于此算法的专用计算单元的一个特殊功能是在每个周期l的开始执行一个新的可变量子值,当剩余值和未知变量形成时,该新量子值将反映在当前周期中将它们向左移动1或2位。通过加或减第一和第二变量差的量子符号,在每次迭代中取值α来进行单元中未知变量的形成。 1.变量归一化的这一特征表示有可能基于整数数据表示来组织计算过程。在单元中算法操作的最后一步,可以考虑最小转换量子的权重,将变量的值恢复为原始实数形式。以FPGA为导向,可以得到所开发算法的硬件和时间资源的比较估计,以及专用计算单元功能有效性的综合比较估计。对于已开发的单元功能算法,本文表明,与特殊算法相比,可以减少一次迭代和整个迭代过程的执行,减少硬件资源的数量,并总体上提高效率。目的计算单元基于简单迭代方法运行。

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