基于GPU的模态分析并行算法  

Parallel algorithms for modal analysis based on GPU

作　　者：朱彬[1] 张宜生[1] 王梁[1] 田晓薇[1] 

机构地区：[1]华中科技大学材料成形与模具技术国家重点实验室,武汉湖北430074

出　　处：《华中科技大学学报（自然科学版）》2012年第5期33-36,共4页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家重点基础研究发展计划资助项目(2010CB630803)

摘　　要：开发了基于图形处理器(GPU)的Cholesky分解并行算法,应用于模态计算程序中,对计算进行加速.算例测试表明该算法相对串行算法计算性能大幅提升,且加速比随矩阵阶数增加而增加,与串行程序相比加速比可达到19.6,此时GPU浮点运算能力达到298Gflops.GPU程序固有频率计算结果与Abaqus计算结果的误差在2%以内,具有足够的计算精度.The Cholesky decomposition parallel algorithm based on graphic processing unit（GPU） was proposed to accelerate the computing speed in the modal calculation program.The example test shows that this algorithm significantly increases computing performance compared with the serial algorithm.The speedup increases with the rise of the matrix order,the value of which can reach to 19.6 compared with the serial program.At the same time,GPU floating point performance reaches to 298 Gflops.At last the error of the natural frequency between the result calculated by the program made by GPU and that of Abaqus is less than 2%.

关 键 词：子空间迭代 图形处理器 模态 并行计算 有限元 

分 类 号：TG241[金属学及工艺—铸造] O327[理学—一般力学与力学基础]