In optimistic simulations, checkpointing techniques are often used to reduce the overhead caused by state saving. In this paper, we propose event reconstruction as a technique with which to reduce the overhead caused by event saving, and compare its memory consumption and execution time to the results obtained by dynamic checkpointing. As the name implies, event reconstruction reconstructs input events and anti-events from the differences between adjacent states, and does not save input events in the event queue. For simulations with fine event granularity and small state size, such as the logic simulation of VLSI circuitry, event reconstruction can yield an improvement in execution time as well as a significant reduction in memory utilization when compared to dynamic checkpointing. Moreover, this technique facilitates load migration because only the state queue needs to be moved from one processor to another.
{"title":"Event reconstruction in time warp","authors":"Lijun Li, C. Tropper","doi":"10.1145/1013329.1013337","DOIUrl":"https://doi.org/10.1145/1013329.1013337","url":null,"abstract":"In optimistic simulations, checkpointing techniques are often used to reduce the overhead caused by state saving. In this paper, we propose event reconstruction as a technique with which to reduce the overhead caused by event saving, and compare its memory consumption and execution time to the results obtained by dynamic checkpointing. As the name implies, event reconstruction reconstructs input events and anti-events from the differences between adjacent states, and does not save input events in the event queue. For simulations with fine event granularity and small state size, such as the logic simulation of VLSI circuitry, event reconstruction can yield an improvement in execution time as well as a significant reduction in memory utilization when compared to dynamic checkpointing. Moreover, this technique facilitates load migration because only the state queue needs to be moved from one processor to another.","PeriodicalId":326595,"journal":{"name":"18th Workshop on Parallel and Distributed Simulation, 2004. PADS 2004.","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128058892","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper advocates the use of a formal framework for analyzing simulation performance. Simulation performance is characterized based on the three simulation development process boundaries: physical system, simulation model, and simulator implementation. Firstly, we formalize simulation event ordering using partially ordered set theory. A simulator implements a simulation event ordering, and incurs implementation overheads when enforcing event ordering at runtime. Secondly, we apply our formalism to extract and formalize the simulation event orderings of both sequential and parallel simulations. Thirdly, we propose the relation stricter and a measure called strictness for comparing and quantifying the degree of event dependency of simulation event orderings respectively.
{"title":"Formalization and strictness of simulation event orderings","authors":"Y. M. Teo, B. Onggo","doi":"10.1145/1013329.1013345","DOIUrl":"https://doi.org/10.1145/1013329.1013345","url":null,"abstract":"This paper advocates the use of a formal framework for analyzing simulation performance. Simulation performance is characterized based on the three simulation development process boundaries: physical system, simulation model, and simulator implementation. Firstly, we formalize simulation event ordering using partially ordered set theory. A simulator implements a simulation event ordering, and incurs implementation overheads when enforcing event ordering at runtime. Secondly, we apply our formalism to extract and formalize the simulation event orderings of both sequential and parallel simulations. Thirdly, we propose the relation stricter and a measure called strictness for comparing and quantifying the degree of event dependency of simulation event orderings respectively.","PeriodicalId":326595,"journal":{"name":"18th Workshop on Parallel and Distributed Simulation, 2004. PADS 2004.","volume":"2014 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121316946","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：