T. Rauber and G. Runger. PVM and MPI Communication Operations on the IBM SP2: Modeling and Comparison. Proc. of the HPCS'97, 1997. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Scheduling Using Genetic Algorithms - Fissgus (2000) (Correct)
....selects for each particular task the suitable implementation (taken, e.g. from a predefined set of library functions) and the appropriate data distributions, and specifies the data redistributions, if necessary. The algorithm is based on runtime prediction formulas proposed by Rauber and Runger [12], 14] and Dierstein [3] and adapted (Sect. 2.3, 8] for the requirements of the scheduling algorithm. In the following, we describe how the scheduling algorithm works. First, we outline the programming model and describe the elements needed for scheduling in Section 2. Especially, this section ....
.... itself is realized by single to single transfer, which depends on the parameters of the target machine model used: number of processors, execution time for an arithmetic operation, startup time for point to point communication, byte transfer time, bandwidth of the interconnection network, see [2] [12], 8] for details. A composed module consists of tasks which represent basic modules, i.e. modules which are computed in a data parallel manner. The computation time of a basic module can be a measured time, e.g. for already available library functions, or it can be an analyticaly derived ....
T. Rauber and G. Runger. PVM and MPI Communication Operations on the IBM SP2: Modeling and Comparison. Proc. of the HPCS'97, 1997.
Modeling the Communication Behavior on Distributed .. - Heinrich-Litan.. (1998) (Correct)
....fact that parallel algorithms on DMMs may show a complicated runtime behavior caused by communication overhead and load balance. Thus, there is considerable research effort to model the performance of DMMs. This includes modeling the runtimes of communication operations with parametrized formulas [Jo91,HXA96,XH96,HEB96,RR97,FRR97]. In these papers, This work has been supported in part by DFG grant Ra 524 5 and Mo 645 5. communication operators from different communication libraries have been investigated and compared for specific DMMs including the Connection Machine CM 2 [Jo91] the Cray T3D [XH96,HEB96] the IBM ....
....In these papers, This work has been supported in part by DFG grant Ra 524 5 and Mo 645 5. communication operators from different communication libraries have been investigated and compared for specific DMMs including the Connection Machine CM 2 [Jo91] the Cray T3D [XH96,HEB96] the IBM SP2 [XH96,RR97], and the Intel Paragon [XH96,FRR97] The modeling of the execution time of the communication operations can be used in compiler tools, in parallelizing compilers or simply as concise information of the performance behavior for the application programmer. The contribution of this paper is to show ....
[Article contains additional citation context not shown here]
Rauber, Th., and Runger, G.: PVM and MPI Communication Operations on the IBM SP2: Modeling and Comparison. In Proceedings of HPCS 97, 1997.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC