摘要
提出了一种新的并行算法,该算法采用了预处理技术,以牺牲固定大小的存储空间为代价,获得较快的运算速度。通过分析以及模拟实验数据表明,与其他同类并行算法相比并行方式上具有更大的灵活性,数据分布也更简单,便于并行实现,可获得更高的加速比。
The thesis puts forward a new sort of parallel algorithm , which adopts precondditioning technique to obtain quick operational speed at the cost of sacrificing fixed memony space. The analysis and simulating experiment show that it is more flexible at parallel mode compared to others. The data distribution belongs to more flexibility, too. Higher accelerator can be achieved, as well.
出处
《廊坊师范学院学报(自然科学版)》
2008年第4期18-20,共3页
Journal of Langfang Normal University(Natural Science Edition)
关键词
模乘幂
并行
算法
modular multiplication power
parallel
algorithm