| 扩展双精度浮点并行计算:MPI方法 |
| |
| 引用本文： | 杨灿群,杨学军,易会战. 扩展双精度浮点并行计算:MPI方法[J]. 计算机工程与科学, 2010, 32(12): 98. DOI: 10.3969/j.issn.1007-130X.2010.12.026 |
| |
| 作者姓名： | 杨灿群 杨学军 易会战 |
| |
| 基金项目： | 国家自然科学基金重点资助项目 |
| |
| 摘 要： | 双精度浮点并行计算将不能满足高性能计算领域对计算精度的要求,但是目前还没有高性能的超双精度并行计算的解决方法。基于并行编程语言MPI,本文提出了扩展双精度浮点的并行计算实现方法,并且使用精度敏感的圆周率计算BBP算法验证了该方法的正确性和性能。
|
| 关 键 词： | 扩展双精度 MPI BBP算法 |
Double-Extended Precision Floating-Point Parallel Computing: MPI Implementation |
| |
| YANG Can-qun,YANG Xue-jun,YI Hui-zhan. Double-Extended Precision Floating-Point Parallel Computing: MPI Implementation[J]. Computer Engineering & Science, 2010, 32(12): 98. DOI: 10.3969/j.issn.1007-130X.2010.12.026 |
| |
| Authors: | YANG Can-qun YANG Xue-jun YI Hui-zhan |
| |
| Abstract: | Parallel computing in double-precision floating-point arithmetic will be out of accord with the requirements of the computing precision in high performance computing in the future.However,there is no solution to parallel computing in super double-precision arithmetic with high performance.Based on the wildly used parallel programming language MPI,this paper presents a solution of double-extended precision parallel computing.The correctness and performance of this solution is validated by a precision-sensitive BBP algorithm which is used for the computation of Pi. |
| |
| Keywords: | double-extended precision MPI BBP algorithm |
| 本文献已被 万方数据 等数据库收录！ |
| 点击此处可从《计算机工程与科学》浏览原始摘要信息 |
|
点击此处可从《计算机工程与科学》下载免费的PDF全文 |
