{"title":"SafeBTW:一种可伸缩的乐观无风险同步算法","authors":"Yaocheng Zhang, Ge Li","doi":"10.1109/PADS.2012.39","DOIUrl":null,"url":null,"abstract":"A new optimistic synchronization algorithm for Parallel Discrete Event Simulation (PDES) called Safe BTW is proposed in this paper. This new algorithm eliminates risky event processing in the Time Warp processing stage of the original BTW algorithm and is founded on a concept called \"safe causal relation\". In our new algorithm, the length of any chained rollback operations is limited to at most one and cascading rollback effects found in the original BTW algorithm can be completely avoided. Performance tests of Time Warp, Breathing Time Bucket, Breathing Time Warp and the new Safe BTW algorithms using PHOLD model show that the new algorithm obtains better speedups as the computation scales up. Its effective event processing time as a percentage of the total event processing time is higher than that of Time Warp and Breathing Time Warp, proving that excessive rollback operations are successfully controlled. And the extra synchronization overhead it introduces is still much lower than that of Breathing Time Bucket itself and is within tolerable and stable level as the computation scales up.","PeriodicalId":299627,"journal":{"name":"2012 ACM/IEEE/SCS 26th Workshop on Principles of Advanced and Distributed Simulation","volume":"88 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2012-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"SafeBTW: A Scalable Optimistic Yet Non-risky Synchronization Algorithm\",\"authors\":\"Yaocheng Zhang, Ge Li\",\"doi\":\"10.1109/PADS.2012.39\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"A new optimistic synchronization algorithm for Parallel Discrete Event Simulation (PDES) called Safe BTW is proposed in this paper. This new algorithm eliminates risky event processing in the Time Warp processing stage of the original BTW algorithm and is founded on a concept called \\\"safe causal relation\\\". In our new algorithm, the length of any chained rollback operations is limited to at most one and cascading rollback effects found in the original BTW algorithm can be completely avoided. Performance tests of Time Warp, Breathing Time Bucket, Breathing Time Warp and the new Safe BTW algorithms using PHOLD model show that the new algorithm obtains better speedups as the computation scales up. Its effective event processing time as a percentage of the total event processing time is higher than that of Time Warp and Breathing Time Warp, proving that excessive rollback operations are successfully controlled. And the extra synchronization overhead it introduces is still much lower than that of Breathing Time Bucket itself and is within tolerable and stable level as the computation scales up.\",\"PeriodicalId\":299627,\"journal\":{\"name\":\"2012 ACM/IEEE/SCS 26th Workshop on Principles of Advanced and Distributed Simulation\",\"volume\":\"88 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2012-07-15\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2012 ACM/IEEE/SCS 26th Workshop on Principles of Advanced and Distributed Simulation\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/PADS.2012.39\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2012 ACM/IEEE/SCS 26th Workshop on Principles of Advanced and Distributed Simulation","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/PADS.2012.39","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
摘要
提出了一种新的并行离散事件模拟(PDES)的乐观同步算法——Safe BTW。该算法基于“安全因果关系”的概念,消除了原BTW算法中时间扭曲处理阶段的风险事件处理。在我们的新算法中,任何链式回滚操作的长度都被限制在最多一次,并且可以完全避免原BTW算法中存在的级联回滚效应。对Time Warp、Breathing Time Bucket、Breathing Time Warp和基于hold模型的新的Safe BTW算法的性能测试表明,随着计算量的扩大,新算法获得了更好的加速。其有效事件处理时间占总事件处理时间的百分比高于时间扭曲和呼吸时间扭曲,证明成功控制了过多的回滚操作。它引入的额外同步开销仍然比Breathing Time Bucket本身要低得多,并且随着计算的扩展,它在可容忍和稳定的水平内。
SafeBTW: A Scalable Optimistic Yet Non-risky Synchronization Algorithm
A new optimistic synchronization algorithm for Parallel Discrete Event Simulation (PDES) called Safe BTW is proposed in this paper. This new algorithm eliminates risky event processing in the Time Warp processing stage of the original BTW algorithm and is founded on a concept called "safe causal relation". In our new algorithm, the length of any chained rollback operations is limited to at most one and cascading rollback effects found in the original BTW algorithm can be completely avoided. Performance tests of Time Warp, Breathing Time Bucket, Breathing Time Warp and the new Safe BTW algorithms using PHOLD model show that the new algorithm obtains better speedups as the computation scales up. Its effective event processing time as a percentage of the total event processing time is higher than that of Time Warp and Breathing Time Warp, proving that excessive rollback operations are successfully controlled. And the extra synchronization overhead it introduces is still much lower than that of Breathing Time Bucket itself and is within tolerable and stable level as the computation scales up.