The application of parallel and distributedsimulation techniques is often limited by the amount of parallelism available in the model. This holds true for large-scale cell-biological simulations, afield that has emer...
详细信息
ISBN:
(纸本)9780769531595
The application of parallel and distributedsimulation techniques is often limited by the amount of parallelism available in the model. This holds true for large-scale cell-biological simulations, afield that has emerged as data and knowledge concerning these systems increases and biologists call for tools to guide wet-lab experimentation. A promising approach to exploit parallelism in this domain is the integration of spatial aspects, which are often crucial to a model's validity. We describe an optimistic, parallel and distributed variant of the Next-Subvolume Method (NSM), a method that augments the well-known Gillespie Stochastic simulation Algorithm (SSA) with spatial features. We discuss requirements imposed by this application on a parallel discrete event simulation engine to achieve efficient execution. First results of combining NSM and the grid-inspired simulation system AURORA are shown.
Imperfect virtual-to-physical page mappings cause excess conflict misses in direct-mapped CPU caches. Perfect placement minimizes the conflict miss rate and is useful for determining an upper bound on direct-mapped ca...
详细信息
Imperfect virtual-to-physical page mappings cause excess conflict misses in direct-mapped CPU caches. Perfect placement minimizes the conflict miss rate and is useful for determining an upper bound on direct-mapped cache performance. In this paper, we introduce a polynomial-time alternative to perfect placement called stochastic page placement. Our method combines a general optimization technique with trace-driven simulation to find a local minimum of the perfect page placement solution space. Two stochastic placement policies are presented: GA placement uses a genetic algorithm to compute virtual page mappings, and SA placement uses simulated annealing. parallel distributed simulation algorithms for perfect placement and GA placement are also presented. For our workloads, GA placement generally outperforms SA placement, but other careful placement policies that exploit workload properties usually outperform stochastic placement.
Research in the use of DEVS (Discrete Event System Specification) based representation of large models on massively parallel platforms is summarized here. We show that parallel DEVS-representation of large scale model...
详细信息
ISBN:
(纸本)0819424986
Research in the use of DEVS (Discrete Event System Specification) based representation of large models on massively parallel platforms is summarized here. We show that parallel DEVS-representation of large scale models can achieve several orders of magnitude speedup. When mapped onto distributed memory multicomputer systems, additional speedup is obtained. An example of a watershed simulation is presented which has been executed in the DEVS-C++/MPI environment working on the CM-5 and IBM SP2 massively parallel systems.
Software rejuvenation is a technique to prevent aging of running software. Recently, benefits of software rejuvenation in high-performance computing (HPC) areas have been remarked. In this paper, we develop a simple r...
详细信息
ISBN:
(纸本)9781509036011
Software rejuvenation is a technique to prevent aging of running software. Recently, benefits of software rejuvenation in high-performance computing (HPC) areas have been remarked. In this paper, we develop a simple rejuvenation technique for time warp-based parallel and distributedsimulation systems. We also present the analytical model to evaluate the reliability of the system which employs our rejuvenation scheme. Furthermore, numerical examples of a couple of parameters of the analytical model are obtained from an experimental implementation of time warp simulator.
parallel discrete event simulation (PDES) have shown to be an useful paradigm for simulating complex and large-scale models. An individual-oriented approach allows modelers capture complex emerging global behaviors ge...
详细信息
parallel discrete event simulation (PDES) have shown to be an useful paradigm for simulating complex and large-scale models. An individual-oriented approach allows modelers capture complex emerging global behaviors generated by simple local interaction, like observed in self-organized systems. Usually, this type of simulations are highly expensive in terms of computing and communications. One one hand, we can reduce the computing involved in individual interactions by means of developing a robust partitioning method. On the other hand, we have to be able to efficiently handle a huge number of individuals interacting with other individuals stored in memory of remote processors. In this work we will analyze and compare three communication strategies: synchronous and asynchronous message passing (via MPI) and bulk-synchronous parallel (BSP) for our distributed cluster-based individual-oriented fish school simulator. In this type of simulations, the main contributions of our work are: a) we showed that distributed time-driven simulations do not always improve the performance when using synchronous communication strategies, b) we show asynchronous communications strategies are more efficient. In addition, we have verified that the bulk-synchronous parallel method is a scalable.
暂无评论