栅元有效共振积分的CUDA算法设计与实现  

Design and implementation of CUDA algorithm for calculating effective resonance integral

作　　者：任成磊[1] 蒲鹏[2] 韩定定[1] 

机构地区：[1]华东师范大学多维度信息处理重点实验室,上海200241 [2]华东师范大学信息科学技术学院,上海200241

出　　处：《计算机工程与科学》2016年第2期224-230,共7页Computer Engineering & Science

摘　　要：核反应堆中需要实时精确地计算堆芯和增殖材料的有效共振积分或群截面来实现反应堆的安全控制。整个计算过程因为涉及大量的积分运算和庞大的核素截面数据,采用常规的计算方法,计算时耗相当大。基于统一计算设备架构(CUDA)平台,利用图形处理器(GPU)的计算能力,对整个计算过程进行并行化分解,多线程同时运算,大幅度提升计算速度,降低时耗。实验结果表明,在GPU上并行计算所得结果与原始数据没有明显差异,且加速效果显著。Nucleon reactors need real-time accurate calculation of effective resonance integral or group-averaged effective cross section to achieve the safety control of reactors. Because the whole calculation process involves a lot of integral operations and huge section data, the time consumption of conventional calculation methods is considerably huge . Based on the compute unified device architecture （CUDA） platform, the whole calculation process with multithreaded operations at the same time can greatly improve the computing speed and reduce the time consumption with the help of the computing power of the graphic processing unit （GPU）. Experimental results show that the parallel computing results on the GPUs have no significant difference from the original data, and the proposal improves the speedup significantly.

关 键 词：慢化源递推公式 有效共振积分 CUDA 并行计算 GPU 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]