摘要
本文提出了基于事务内存的分布式编程框架,能够有效地减少由事务内存带来的网络带宽消耗,并且降低事务内存中事务操作的通信延迟.其中,重点设计实现了一个新的事物内存缓存数据一致性维护机制.测试结果表明,使用该新型基于事物内存的分布式编程框架实现的分布式计算任务具有简单的设计模型和较高的效率,有效的减少了系统网络带宽需求,在小规模集群中相对于现有的分布式事务内存系统有30%的性能提升.
In this paper, we propose a generalized and scalable distributed programming environment based on transactional memory, which consumes low cost of bandwidth and reduce the latency of transactional operations. In the scheme, a cache coherency scheme has been proposed. Experimentations show that the new scheme has good performance, and has 30%improvement, compared with traditional transactional memory system.
出处
《微电子学与计算机》
CSCD
北大核心
2013年第3期59-62,共4页
Microelectronics & Computer
基金
中央高校基本科研业务费专项资金(XUFL110232)
关键词
分布式
编程框架
事物内存
缓存一致性
distributed
programming environment
transactional memory
cache coherency
