拉普拉斯生长的GPU实现与硬件加速

摘要

编写实现了六方点阵中拉普拉斯模型的随机行走雪花生长的CUDA(Compute Unified Device Architecture)程序,分析了GPU (Graphic Processing Unit)随机行走计算效率,对比了拉普拉斯模型随机行走雪花生长的GPU与CPU加速特性,给出了该模型随机计算环境尺寸增加的用时增长曲线.%A 2D hep grid Laplace random walk model program based on CUDA (Compute Unified Device Architecture) is coded and tested. The computing efficiency result compared to the corresponding CPU (Central Processing Unit) is given and comparing diagrams are presented.