{"title":"Session details: Paper Session 8 Miscellaneous","authors":"Jared S. Ivey","doi":"10.1145/3254057","DOIUrl":"https://doi.org/10.1145/3254057","url":null,"abstract":"","PeriodicalId":341026,"journal":{"name":"Proceedings of the 2017 ACM SIGSIM Conference on Principles of Advanced Discrete Simulation","volume":"126 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116567660","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}
In this keynote talk, we discuss an extended Belief-Desire-Intention (BDI) framework for human decision making and planning, whose sub-modules have been developed using Bayesian belief network (BBN), Decision-Field-Theory (DFT), and a probabilistic depth first search (DFS) technique. A key novelty of the proposed approach is its ability to represent both the human decision-making as well as decision-planning functions in a coherent framework. In this talk, the proposed framework is illustrated and demonstrated for human's evacuation behaviors under a terrorist bomb attack situation. To mimic realistic human decision-planning and decision-making behaviors, attributes of the extended BDI framework are calibrated from the human-in-the-loop experiments conducted in the Cave Automatic Virtual Environment (CAVE) available at The University of Arizona. A crowd simulation is then constructed, where individual human behaviors are modeled based on what was learned from the CAVE experiments. In this work, the simulated environment (e.g. streets and buildings) and humans conforming to the extended BDI framework are implemented in AnyLogic® agent-based simulation software, where each human entity calls external Netica BBN software to perform its perceptual processing function and Soar software to perform its real-time planning and decision-execution functions. The constructed crowd simulation is used to evaluate the impact of several factors (e.g. number of policemen, information sharing via speakers/mobile phones) on the evacuation performance. Finally, we briefly discuss other applications (e.g. driver's behaviors) and research extensions (e.g. human learning/forgetting and human interactions) for the extended BDI framework.
{"title":"An Integrated Human Decision Making Model under Extended Belief-Desire-Intention Framework","authors":"Y. Son","doi":"10.1145/3064911.3064936","DOIUrl":"https://doi.org/10.1145/3064911.3064936","url":null,"abstract":"In this keynote talk, we discuss an extended Belief-Desire-Intention (BDI) framework for human decision making and planning, whose sub-modules have been developed using Bayesian belief network (BBN), Decision-Field-Theory (DFT), and a probabilistic depth first search (DFS) technique. A key novelty of the proposed approach is its ability to represent both the human decision-making as well as decision-planning functions in a coherent framework. In this talk, the proposed framework is illustrated and demonstrated for human's evacuation behaviors under a terrorist bomb attack situation. To mimic realistic human decision-planning and decision-making behaviors, attributes of the extended BDI framework are calibrated from the human-in-the-loop experiments conducted in the Cave Automatic Virtual Environment (CAVE) available at The University of Arizona. A crowd simulation is then constructed, where individual human behaviors are modeled based on what was learned from the CAVE experiments. In this work, the simulated environment (e.g. streets and buildings) and humans conforming to the extended BDI framework are implemented in AnyLogic® agent-based simulation software, where each human entity calls external Netica BBN software to perform its perceptual processing function and Soar software to perform its real-time planning and decision-execution functions. The constructed crowd simulation is used to evaluate the impact of several factors (e.g. number of policemen, information sharing via speakers/mobile phones) on the evacuation performance. Finally, we briefly discuss other applications (e.g. driver's behaviors) and research extensions (e.g. human learning/forgetting and human interactions) for the extended BDI framework.","PeriodicalId":341026,"journal":{"name":"Proceedings of the 2017 ACM SIGSIM Conference on Principles of Advanced Discrete Simulation","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129902107","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}
Davide Cingolani, Alessandro Pellegrini, M. Schordan, F. Quaglia, D. Jefferson
State recoverability is a crucial aspect of speculative Time Warp-based Parallel Discrete Event Simulation. In the literature, we can identify three major classes of techniques to support the correct restoration of a previous simulation state upon the execution of a rollback operation: state checkpointing/restore, manual reverse computation and automatic reverse computation. The latter class has been recently supported by relying either on binary code instrumentation or on source-to-source code transformation. Nevertheless, both solutions are not intrinsically meant to support a reversible execution of third-party shared libraries, which can be pretty useful when implementing complex simulation models. In this paper, we present an architectural solution (realized as a static C library) which allows to transparently instrument at runtime any third party shared library, with no need for any modification to the model's code. We also present a preliminary experimental evaluation, based on the integration of our library with the ROOT-Sim simulation engine.
{"title":"Dealing with Reversibility of Shared Libraries in PDES","authors":"Davide Cingolani, Alessandro Pellegrini, M. Schordan, F. Quaglia, D. Jefferson","doi":"10.1145/3064911.3064927","DOIUrl":"https://doi.org/10.1145/3064911.3064927","url":null,"abstract":"State recoverability is a crucial aspect of speculative Time Warp-based Parallel Discrete Event Simulation. In the literature, we can identify three major classes of techniques to support the correct restoration of a previous simulation state upon the execution of a rollback operation: state checkpointing/restore, manual reverse computation and automatic reverse computation. The latter class has been recently supported by relying either on binary code instrumentation or on source-to-source code transformation. Nevertheless, both solutions are not intrinsically meant to support a reversible execution of third-party shared libraries, which can be pretty useful when implementing complex simulation models. In this paper, we present an architectural solution (realized as a static C library) which allows to transparently instrument at runtime any third party shared library, with no need for any modification to the model's code. We also present a preliminary experimental evaluation, based on the integration of our library with the ROOT-Sim simulation engine.","PeriodicalId":341026,"journal":{"name":"Proceedings of the 2017 ACM SIGSIM Conference on Principles of Advanced Discrete Simulation","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125019071","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}
Yulin Wu, Xiangting Hou, Wen Jun Tan, Zengxiang Li, Wentong Cai
Social contact network (SCN) models the contacts between people by their daily activities. It can be formalized by an agent-to-location bipartite graph. The simulations over SCN are employed to study the complex social dynamics such as information propagation and disease spread among large-scale population. A challenge to the simulation is the skewed degree distribution of SCN, which contains a few hub locations with large numbers of visitors. The skewed degree distribution can cause load imbalance for parallel simulation and greatly degrade the execution performance. This paper proposes an approach which decomposes hub locations into small splits. Thus, SCN can be partitioned with better balanced workloads and multiple splits are able to run in parallel. Based on the pattern of information transmission between agents, we duplicate necessary data among splits to ensure the correctness of simulation. Furthermore, we enhance the parallel algorithm of SCN simulation to reduce the additional overhead from communication between splits. Finally, we build an experiment with epidemic simulation on an open dataset. The experimental results demonstrate that our approach achieves 14~35% performance improvement compared with the partitioning method without decomposition of hub locations.
{"title":"Efficient Parallel Simulation over Social Contact Network with Skewed Degree Distribution","authors":"Yulin Wu, Xiangting Hou, Wen Jun Tan, Zengxiang Li, Wentong Cai","doi":"10.1145/3064911.3064934","DOIUrl":"https://doi.org/10.1145/3064911.3064934","url":null,"abstract":"Social contact network (SCN) models the contacts between people by their daily activities. It can be formalized by an agent-to-location bipartite graph. The simulations over SCN are employed to study the complex social dynamics such as information propagation and disease spread among large-scale population. A challenge to the simulation is the skewed degree distribution of SCN, which contains a few hub locations with large numbers of visitors. The skewed degree distribution can cause load imbalance for parallel simulation and greatly degrade the execution performance. This paper proposes an approach which decomposes hub locations into small splits. Thus, SCN can be partitioned with better balanced workloads and multiple splits are able to run in parallel. Based on the pattern of information transmission between agents, we duplicate necessary data among splits to ensure the correctness of simulation. Furthermore, we enhance the parallel algorithm of SCN simulation to reduce the additional overhead from communication between splits. Finally, we build an experiment with epidemic simulation on an open dataset. The experimental results demonstrate that our approach achieves 14~35% performance improvement compared with the partitioning method without decomposition of hub locations.","PeriodicalId":341026,"journal":{"name":"Proceedings of the 2017 ACM SIGSIM Conference on Principles of Advanced Discrete Simulation","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129589111","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}
Hajj is one of the largest mass gatherings where Muslims from all over the world gather in Makah each year for pilgrimage. A mass assembly of such scale bears a huge risk of disaster either natural or man-made. In the past few years, thousands of casualties have occurred while performing different Hajj rituals, especially during the Circumambulation of Kaba (Tawaf) due to stampede or chaos. During such calamitous situations, an appropriate evacuation strategy can help resolve the problem and mitigate further risk of causalities. It is however a daunting research problem to identify an optimal course of action based on several constraints. Modeling and analyzing such a problem of real-time and spatially explicit complexity requires a microscale crowd simulation and analysis framework. Which not only allows the modeler to express the spatial dimensions and features of the environment in real scale, but also provides modalities to capture complex crowd behaviors. In this paper, we propose an Agent-based Crowd Simulation & Analysis framework that incorporates the use of Anylogic Pedestrian library and integrates/interoperate Anylogic Simulation environment with the external modules for optimization and analysis. Hence provides a runtime environment for analyzing complex situations, e.g., emergency evacuation strategies. The key features of the proposed framework include: (i) Ability to model large crowd in a spatially explicit environment at real-scale; (ii) Simulation of complex crowd behavior such as emergency evacuation; (iii) Interoperability of optimization and analysis modules with simulation runtime for evaluating evacuation strategies. We present a case study of Hajj scenario as a proof of concept and a test bed for identifying and evaluating optimal strategies for crowd evacuation
{"title":"Analyzing Emergency Evacuation Strategies for Mass Gatherings using Crowd Simulation And Analysis framework: Hajj Scenario","authors":"Imran Mahmood, Muhammad Haris, H. Sarjoughian","doi":"10.1145/3064911.3064924","DOIUrl":"https://doi.org/10.1145/3064911.3064924","url":null,"abstract":"Hajj is one of the largest mass gatherings where Muslims from all over the world gather in Makah each year for pilgrimage. A mass assembly of such scale bears a huge risk of disaster either natural or man-made. In the past few years, thousands of casualties have occurred while performing different Hajj rituals, especially during the Circumambulation of Kaba (Tawaf) due to stampede or chaos. During such calamitous situations, an appropriate evacuation strategy can help resolve the problem and mitigate further risk of causalities. It is however a daunting research problem to identify an optimal course of action based on several constraints. Modeling and analyzing such a problem of real-time and spatially explicit complexity requires a microscale crowd simulation and analysis framework. Which not only allows the modeler to express the spatial dimensions and features of the environment in real scale, but also provides modalities to capture complex crowd behaviors. In this paper, we propose an Agent-based Crowd Simulation & Analysis framework that incorporates the use of Anylogic Pedestrian library and integrates/interoperate Anylogic Simulation environment with the external modules for optimization and analysis. Hence provides a runtime environment for analyzing complex situations, e.g., emergency evacuation strategies. The key features of the proposed framework include: (i) Ability to model large crowd in a spatially explicit environment at real-scale; (ii) Simulation of complex crowd behavior such as emergency evacuation; (iii) Interoperability of optimization and analysis modules with simulation runtime for evaluating evacuation strategies. We present a case study of Hajj scenario as a proof of concept and a test bed for identifying and evaluating optimal strategies for crowd evacuation","PeriodicalId":341026,"journal":{"name":"Proceedings of the 2017 ACM SIGSIM Conference on Principles of Advanced Discrete Simulation","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116085532","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}
The switch from gasoline-powered vehicles to electric vehicles (EVs) is an important step to reduce greenhouse gas emissions. To this end, many European countries announced EV stock targets, e.g., Germany aims to have one million EVs on the roads by 2020. To achieve these goals and to handle the range limitation of EVs, a widespread publicly accessible charging infrastructure is needed. This paper provides a dynamic spatial and temporal simulation model for the building of charging infrastructure on a municipality scale. We evaluate empirical data about the timely utilization of different charging stations in the German federal state of Bavaria. This data is used to derive empirical models for the start time and duration of charging events as well as the popularity of charging stations. We develop a lightweight discrete event simulation model which can be used to investigate different expansion strategies, e.g., based on the load of charging stations or the number of successful and failed charging attempts. We show the applicability of our model using the German federal state of Bavaria as a use case.
{"title":"Spatial and Temporal Charging Infrastructure Planning Using Discrete Event Simulation","authors":"M. Pruckner, R. German, D. Eckhoff","doi":"10.1145/3064911.3064919","DOIUrl":"https://doi.org/10.1145/3064911.3064919","url":null,"abstract":"The switch from gasoline-powered vehicles to electric vehicles (EVs) is an important step to reduce greenhouse gas emissions. To this end, many European countries announced EV stock targets, e.g., Germany aims to have one million EVs on the roads by 2020. To achieve these goals and to handle the range limitation of EVs, a widespread publicly accessible charging infrastructure is needed. This paper provides a dynamic spatial and temporal simulation model for the building of charging infrastructure on a municipality scale. We evaluate empirical data about the timely utilization of different charging stations in the German federal state of Bavaria. This data is used to derive empirical models for the start time and duration of charging events as well as the popularity of charging stations. We develop a lightweight discrete event simulation model which can be used to investigate different expansion strategies, e.g., based on the load of charging stations or the number of successful and failed charging attempts. We show the applicability of our model using the German federal state of Bavaria as a use case.","PeriodicalId":341026,"journal":{"name":"Proceedings of the 2017 ACM SIGSIM Conference on Principles of Advanced Discrete Simulation","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125371847","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}
Software-defined networking (SDN) technology promises centralized and rapid network provisioning, holistic management, low operational cost, and improved network visibility. Researchers have developed multiple SDN simulation and emulation platforms to expedite the adoption of many emerging SDN-based applications to production systems. However, the scalability of those platforms is often limited by the underlying physical hardware resources, which inevitably affects the simulation fidelity in large-scale network settings. In this paper, we present a model abstraction technique that effectively transforms the network devices in an SDN-based network to one virtualized switch model. While significantly reducing the model execution time and enabling the real-time simulation capability, our abstracted model also preserves the end-to-end forwarding behavior of the original network. To achieve this, we first classify packets with the same forwarding behavior into smaller and disjoint Equivalence Classes (ECes) by analyzing the OpenFlow rules installed on the SDN devices. We then create a graph model representing the forwarding behavior of each EC. By traversing those graphs, we finally construct the rules of the big-switch model to effectively preserve the original network's end-to-end forwarding behavior. Experimental results demonstrate that the network forwarding logic equivalence is well preserved between the abstracted model and the original SDN network. The model abstraction process is fast, e.g., 3.15 seconds to transform a medium-scale tree network consisting of 53,260 rules. The big-switch model is able to speed up the simulation by 4.3 times in average and up to 6.69 times among our evaluation experiments.
{"title":"Simulation of a Software-Defined Network as One Big Switch","authors":"Jiaqi Yan, Xin Liu, Dong Jin","doi":"10.1145/3064911.3064918","DOIUrl":"https://doi.org/10.1145/3064911.3064918","url":null,"abstract":"Software-defined networking (SDN) technology promises centralized and rapid network provisioning, holistic management, low operational cost, and improved network visibility. Researchers have developed multiple SDN simulation and emulation platforms to expedite the adoption of many emerging SDN-based applications to production systems. However, the scalability of those platforms is often limited by the underlying physical hardware resources, which inevitably affects the simulation fidelity in large-scale network settings. In this paper, we present a model abstraction technique that effectively transforms the network devices in an SDN-based network to one virtualized switch model. While significantly reducing the model execution time and enabling the real-time simulation capability, our abstracted model also preserves the end-to-end forwarding behavior of the original network. To achieve this, we first classify packets with the same forwarding behavior into smaller and disjoint Equivalence Classes (ECes) by analyzing the OpenFlow rules installed on the SDN devices. We then create a graph model representing the forwarding behavior of each EC. By traversing those graphs, we finally construct the rules of the big-switch model to effectively preserve the original network's end-to-end forwarding behavior. Experimental results demonstrate that the network forwarding logic equivalence is well preserved between the abstracted model and the original SDN network. The model abstraction process is fast, e.g., 3.15 seconds to transform a medium-scale tree network consisting of 53,260 rules. The big-switch model is able to speed up the simulation by 4.3 times in average and up to 6.69 times among our evaluation experiments.","PeriodicalId":341026,"journal":{"name":"Proceedings of the 2017 ACM SIGSIM Conference on Principles of Advanced Discrete Simulation","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129549521","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}
Jonatan Lindén, Pavol Bauer, Stefan Engblom, B. Jonsson
We present a new efficient approach to the parallelization of discrete event simulators for multicore computers, which is based on exposing and disseminating essential information between processors. We aim specifically at simulation models with a spatial structure, where time intervals between successive events are highly variable and without lower bounds. In Parallel Discrete Event Simulation (PDES), the model is distributed onto parallel processes. A key challenge in PDES is that each process must continuously decide when to pause its local simulation in order to reduce the risk of expensive rollbacks caused by future "delayed"' incoming events from other processes. A process could make such decisions optimally if it would know the timestamps of future incoming events. Unfortunately, this information is often not available in PDES algorithms. We present an approach to designing efficient PDES algorithms, in which an existing natural parallelization of PDES is restructured in order to expose and disseminate more precise information about future incoming events to each LP. We have implemented our approach in a parallel simulator for spatially extended Markovian processes, intended for simulating, e.g., chemical reactions, biological and epidemiological processes. On 32 cores, our implementation exhibits speedup that significantly outweighs the overhead incurred by the refinement. We also show that our resulting simulator is superior in performance to existing simulators for comparable models, achieving for 32 cores an average speedup of 20 relative to an efficient sequential implementation.
{"title":"Exposing Inter-Process Information for Efficient Parallel Discrete Event Simulation of Spatial Stochastic Systems","authors":"Jonatan Lindén, Pavol Bauer, Stefan Engblom, B. Jonsson","doi":"10.1145/3064911.3064916","DOIUrl":"https://doi.org/10.1145/3064911.3064916","url":null,"abstract":"We present a new efficient approach to the parallelization of discrete event simulators for multicore computers, which is based on exposing and disseminating essential information between processors. We aim specifically at simulation models with a spatial structure, where time intervals between successive events are highly variable and without lower bounds. In Parallel Discrete Event Simulation (PDES), the model is distributed onto parallel processes. A key challenge in PDES is that each process must continuously decide when to pause its local simulation in order to reduce the risk of expensive rollbacks caused by future \"delayed\"' incoming events from other processes. A process could make such decisions optimally if it would know the timestamps of future incoming events. Unfortunately, this information is often not available in PDES algorithms. We present an approach to designing efficient PDES algorithms, in which an existing natural parallelization of PDES is restructured in order to expose and disseminate more precise information about future incoming events to each LP. We have implemented our approach in a parallel simulator for spatially extended Markovian processes, intended for simulating, e.g., chemical reactions, biological and epidemiological processes. On 32 cores, our implementation exhibits speedup that significantly outweighs the overhead incurred by the refinement. We also show that our resulting simulator is superior in performance to existing simulators for comparable models, achieving for 32 cores an average speedup of 20 relative to an efficient sequential implementation.","PeriodicalId":341026,"journal":{"name":"Proceedings of the 2017 ACM SIGSIM Conference on Principles of Advanced Discrete Simulation","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128537559","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}
R. Fujimoto, M. Hunter, A. Biswas, Mark Jackson, Sabra A. Neal
Energy and power consumption have become important concerns for many computing systems ranging from embedded and mobile systems operating on battery-powered devices to high performance and cloud computing applications running on supercomputers and in data centers. To date, only a limited amount of work has considered power consumption in parallel and distributed simulations. A variety of options to analyze and explore power consumption in distributed simulations are discussed. These options range from design decisions in developing the simulation model to selection of algorithms in distributed simulation middleware to exploitation of hardware techniques. Work to characterize the power and energy consumed by different elements of parallel and distributed simulation systems are discussed and empirical measurements presented to quantify energy and power use, suggestive of directions for future research in this area.
{"title":"Power Efficient Distributed Simulation","authors":"R. Fujimoto, M. Hunter, A. Biswas, Mark Jackson, Sabra A. Neal","doi":"10.1145/3064911.3069397","DOIUrl":"https://doi.org/10.1145/3064911.3069397","url":null,"abstract":"Energy and power consumption have become important concerns for many computing systems ranging from embedded and mobile systems operating on battery-powered devices to high performance and cloud computing applications running on supercomputers and in data centers. To date, only a limited amount of work has considered power consumption in parallel and distributed simulations. A variety of options to analyze and explore power consumption in distributed simulations are discussed. These options range from design decisions in developing the simulation model to selection of algorithms in distributed simulation middleware to exploitation of hardware techniques. Work to characterize the power and energy consumed by different elements of parallel and distributed simulation systems are discussed and empirical measurements presented to quantify energy and power use, suggestive of directions for future research in this area.","PeriodicalId":341026,"journal":{"name":"Proceedings of the 2017 ACM SIGSIM Conference on Principles of Advanced Discrete Simulation","volume":"73 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131682490","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}
B. Williams, D. Ponomarev, N. Abu-Ghazaleh, P. Wilsey
Performance and scalability of Parallel Discrete Event Simulation (PDES) is often limited by fine-grain communication, especially in execution environments with high communication cost. However, the low cost of on-chip communication in emerging many-core processors offers a promise to substantially alleviate conventional PDES bottlenecks. In this paper, we present a detailed evaluation and characterization of multi-threaded ROSS simulator on Intel's Knights Landing (KNL) processor. KNL is the second generation of the Intel Xeon Phi family of processors offering significant architecture improvements including 64 out-of-order multithreaded cores, sharing of some levels of the cache hierarchy among the cores, fast 2D mesh interconnect network and the ability to reconfigure the processor to support various clustering modes. We analyze the performance and scalability of ROSS simulator on KNL processor under different thread counts, communication patterns, event processing granularities, synchronization periods, thread placement policies, and workload partitioning schemes. We conclude that within a single KNL processor, up to 2X performance improvement can be achieved compared to commodity Xeon multicore processors. We show that in most cases the performance of ROSS scales well with the best results achieved when thread affinity is assigned, CPU cores are evenly loaded, cache sharing is exploited and communication is limited to small clusters of cores.
{"title":"Performance Characterization of Parallel Discrete Event Simulation on Knights Landing Processor","authors":"B. Williams, D. Ponomarev, N. Abu-Ghazaleh, P. Wilsey","doi":"10.1145/3064911.3064929","DOIUrl":"https://doi.org/10.1145/3064911.3064929","url":null,"abstract":"Performance and scalability of Parallel Discrete Event Simulation (PDES) is often limited by fine-grain communication, especially in execution environments with high communication cost. However, the low cost of on-chip communication in emerging many-core processors offers a promise to substantially alleviate conventional PDES bottlenecks. In this paper, we present a detailed evaluation and characterization of multi-threaded ROSS simulator on Intel's Knights Landing (KNL) processor. KNL is the second generation of the Intel Xeon Phi family of processors offering significant architecture improvements including 64 out-of-order multithreaded cores, sharing of some levels of the cache hierarchy among the cores, fast 2D mesh interconnect network and the ability to reconfigure the processor to support various clustering modes. We analyze the performance and scalability of ROSS simulator on KNL processor under different thread counts, communication patterns, event processing granularities, synchronization periods, thread placement policies, and workload partitioning schemes. We conclude that within a single KNL processor, up to 2X performance improvement can be achieved compared to commodity Xeon multicore processors. We show that in most cases the performance of ROSS scales well with the best results achieved when thread affinity is assigned, CPU cores are evenly loaded, cache sharing is exploited and communication is limited to small clusters of cores.","PeriodicalId":341026,"journal":{"name":"Proceedings of the 2017 ACM SIGSIM Conference on Principles of Advanced Discrete Simulation","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128997789","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：