基于并行算法的大数据阶乘算法的时间效率优化分析  

Time Efficiency Optimization Analysis of Big Data Factorial Algorithm Based on Parallel Algorithm

作　　者：高鑫 王世杰 许舒翔 GAO Xin;WANG Shijie;XU Shuxiang(Northwest Branch,State Grid Co.,Ltd.,Xi’an 710048,China;NARI Technology Co.,Ltd.,Nanjing 211106,China)

机构地区：[1]国家电网公司西北分部,陕西西安710048 [2]国电南瑞科技股份有限公司,江苏南京211106

出　　处：《微型电脑应用》2021年第1期168-169,173,共3页Microcomputer Applications

摘　　要：为了提高并行算法计算量,设计了基于并行算法的大数阶乘运算方法,可以在有限的硬件资源条件下根据不同的问题规模为计算过程分配合适的存储空间,并且可以发挥FPGA所具备的并行处理功能,对多核处理器的并行处理过程进行模拟分析。VHDL通过并行语句的模式来构建Ala语句,各权重单元都实施阶乘运算,当权重单元发生溢出的情况时再迭代更新。利用此算法开发得到了可以实现多种功能的阶乘计算器上位机,显著提升了时空效率,有效满足了大数阶乘运算的需求。与现有阶乘算法相比,算法所需的用时最短。在各复杂度下本文算法都相对于传统算法实现了效率的明显提升。In order to improve the parallel algorithm,we design a factorial large numbers of operation method based on the parallel algorithm,which can work under the condition of limited hardware resources according to the different problem size to distribute suitable storage space for calculating process,and can play as an FPGA with parallel processing features,and can implement simulation analysis for the parallel processing of multi-core processor.VHDL constructs Ala statement through the mode of parallel statement.All weight units perform factorial operation and update iteratively when overflow occurs in the weight unit.Using this algorithm,the upper computer of the factorial calculator is developed to realize many functions.Compared with the existing factorial algorithm,the time required by this algorithm is the shortest.Under various complexity,the efficiency of the proposed algorithm is obviously improved compared with that of the traditional algorithm.

关 键 词：阶乘算法 存储空间 并行计算 FPGA 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]