共查询到18条相似文献,搜索用时 93 毫秒
1.
提出一种适用于SMP集群的混合MPI+OpenMP并行编程模型。该模型贴近于SMP集群的体系结构且综合了消息传递和共享内存2种编程模型的优势,能获得较好的性能。讨论该混合模型的实现机制以及MPI消息传递模型的特点。实验结果表明,在一定条件下,该混合并行编程模型是SMP集群的最优选择。 相似文献
2.
基于SMP集群的三维网格多粒度混合并行编程模型 总被引:2,自引:0,他引:2
为提高大规模三维网格并行算法的执行效率,针对SMP集群分布/共享两级内存层次结构的特点,介绍适用于SMP集群混合编程的不同实现方法.对三维网格模型最短路径问题的并行求解提出了多粒度混合并行编程模型,给出了实现该问题的MPI+OpenMP混合并行算法,并在SMP集群上同粗粒度MPI(Message Passing Interface)并行算法做了性能比较.结果表明,采用该多粒度混合并行编程模型具有更好的加速比和运行效率. 相似文献
3.
为了提高分子动力学模拟在对称多处理(SMP)集群上的计算速度,在分子动力学并行方法中引入MPI+TBB的混合并行编程模型。基于该模型,在分子动力学软件LAMMPS中设计并实现混合并行算法,在节点间采用MPI及空间分解技术实施进程级并行,节点内采用TBB及临界区技术实施线程级并行。在SMP集群中的测试表明,该方法在体系较大以及节点数较多时可以明显减少通信时间,使加速比在纯MPI模型上提高45%。结果表明,MPI+TBB混合并行编程模型可促进分子动力学并行模拟且效率明显提升。 相似文献
4.
对采用多核处理器作为SMP集群系统的计算节点的系统上的一种混合编程模型-MPI+OpenMP混合编程模型进行了深入的研究.建立了两个矩阵乘的混合并行算法,在多核集群平台上与纯MPI算法分别进行了实验,并进行了性能方面的比较.试验表明,混合编程具有更好的性能. 相似文献
5.
SMP机群混合编程模型研究 总被引:12,自引:0,他引:12
研究了适用于 SMP机群的混合编程模型 ,并把它划分为 Open MP MPI和 Thread MPI两类 .通过研究指出 ,Open MP MPI优于 Thread MPI.在此基础上 ,重点研究了 Open MP MPI的实现机制、粗粒度和细粒度并行化方法、循环选择、优化措施以及注意事项等 ,得出细粒度并行化的 Open MP MPI是 SMP机群编程模型的一个较好选择的结论 相似文献
6.
7.
主要对并行计算的编程模型进行了研究,包括MPI和OpenMP两种编程模型,同时提出了一种层次化混合编程模型。并以计算π的问题为例,用C语言设计了混合编程模型下的程序,在以多核处理器作为节点的曙光TC5000集群上对三种编程模型下的求π程序进行了实验,同时将实验结果进行了性能分析和比较。结果表明该混合并行算法具有更好的扩展性和加速比。 相似文献
9.
10.
一个数值天气预报模式的并行混合编程模型及其应用 总被引:3,自引:0,他引:3
朱政慧 《数值计算与计算机应用》2005,26(3):198-207
提出了一个适用于数值天气预报模式的并行混合编程模型,归纳出混合并行算法设计的特点.依据该模型,针对国内发展的区域数值天气预报模式,开发基于MPI/OpenMP的并行版本.通过对五种并行方案的并行性能分析比较,发现其结果同抽象模型的分析结论是一致的. 相似文献
11.
12.
13.
This research defines and analyzes a methodology for deriving a performance model for SPMD hybrid parallel applications. Hybrid parallelism combines shared memory and message passing computing models. This work extends the current practice of application performance modelling by development of a methodology for hybrid applications with these procedures.
- Creation of a model based on complexity analysis of an application code and its data structures.
- Enhancement of a static complexity model by dynamic factors to capture execution time phenomena, such as memory hierarchy effects.
- Quantitative analysis of model characteristics and the effects of perturbations in measured parameters.
14.
15.
WAPM:适合广域分布式计算的并行编程模型 总被引:1,自引:0,他引:1
早期的MPI与OpenMP等编程模型由于扩展性限制或并行粒度的差异而不适合于大规模的广域动态Internet环境.提出了一个用于广域网络范围内的并行编程模型(WAPM),为应用的分布式计算的编程提供了一个新的可行解决方案.WAPM由通信库、通信协议和应用编程接口组成,并且具有通用编程、自适应并行、容错性等特点,通过选择合适的编程语言,就可形成一个广域范围内的并行程序设计环境.以分布式计算平台P2HP为工作平台,描述了WAPM分布式计算的实施过程.实验结果表明,WAPM是一个通用的、可行的、性能较好的编程模型. 相似文献
16.
With the advancement of new processor and memory architectures, supercomputers of multicore and multinode architectures have become general tools for large‐scale engineering and scientific simulations. However, the nonuniform latencies between intranode and internode communications on these machines introduce new challenges that need to be addressed in order to achieve optimal performance. In this paper, a novel hybrid solver that is especially designed for supercomputers of multicore and multinode architectures is proposed. The new hybrid solver is characterized by its two‐level parallel computing approach on the basis of the strategies of two‐level partitioning and two‐level condensation. It distinguishes intranode and internode communications to minimize the communication overheads. Moreover, it further reduces the size of interface equation system to improve its convergence rate. Three numerical experiments of structural linear static analysis were conducted on DAWNING‐5000A supercomputer to demonstrate the validity and efficiency of the proposed method. Test results show that the proposed approach was superior in performance compared with the conventional Schur complement method. Copyright © 2014 John Wiley & Sons, Ltd. 相似文献
17.
随着混合异构平台的发展,出现了类型不一的加速设备,如何在混合异构平台中充分利用这些不同类型的设备,以及如何在多个计算设备之间部署深度学习模型,而且训练大型和复杂模型的重要性不断提高.数据并行(DP)是应用最广泛的并行化策略,但是如果数据并行训练中的设备数量不断增加,设备之间的通信开销就会成为瓶颈.此外,每个步骤因设备性... 相似文献
18.
Clusters of symmetric multiprocessors (SMPs) are popular platforms for parallel programming since they provide large computational power for a reasonable price. For irregular application programs with dynamically changing computation and data access behavior, a flexible programming model is needed to achieve efficiency. In this paper we propose Task Pool Teams as a hybrid parallel programming environment to realize irregular algorithms on clusters of SMPs. Task Pool Teams combine task pools on single cluster nodes by an explicit message passing layer. They offer load balance together with multi‐threaded, asynchronous communication. Appropriate communication protocols and task pool implementations are provided and accessible by an easy‐to‐use application programmer interface. As application examples we present a branch and bound algorithm and the hierarchical radiosity algorithm. Copyright © 2006 John Wiley & Sons, Ltd. 相似文献