| T. Rauber and G. Runger. The Compiler TwoL for the Design of Parallel Implementations. In Proc. 4th Int. Conf. on Parallel Architectures and Compilation Techniques (PACT'96), IEEE, pages 292--301, 1996. |
....of the GA operations mutation and crossover is given, and the fitness function, i.e. the function which evaluates the quality of a solution, is described. Section 4 shows experimental results. Finally, Section 5 discusses some future research directions. 2 Model Overview The TwoL model [11], 5] clearly separates task parallelism from data parallelism. This clear separation is used to divide the work between programmer and compiler system. The programmer specifies an algorithm as a module specification and provides a specification of the data parallel basic modules. The module ....
....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 estimation function. We use a runtime estimation model for SPMD programs in message passing style [11]. Investigations for several applications from numerical analysis show that the runtime prediction formulas describe the execution time accurately enough to compare different execution schemes of the same application [14] 8] 3 Genetic Algorithms The usual form of genetic algorithms (GAs) was ....
[Article contains additional citation context not shown here]
T. Rauber and G. Runger. The Compiler TwoL for the Design of Parallel Implementations. In Proc. 4th Int. Conf. on Parallel Architectures and Compilation Techniques, IEEE, pages 282-301, 1996.
No context found.
T. Rauber and G. Runger. The Compiler TwoL for the Design of Parallel Implementations. In Proc. 4th Int. Conf. on Parallel Architectures and Compilation Techniques (PACT'96), IEEE, pages 292--301, 1996.
.... question whether the exploitation of the available degree of task parallelism of a specific application leads to an efficient program for a given DMM and how design decisions like task scheduling and data distribution can be derived from a general specification expressing the potential parallelism [25, 26]. In this paper, we consider the relationship between task and data parallel executions and describe how hierarchically structured task parallelism can automatically be transformed into a corresponding message passing program using a hierarchy of group contexts. Tasks that are not further ....
T. Rauber and G. Runger. The Compiler TwoL for the Design of Parallel Implementations. In Proc. 4th Int. Conf. on Parallel Architectures and Compilation Techniques, IEEE, pages 282--301, 1996.
....multitask parallelism data parallelism cost information group information data distribution type Figure 1. Overview of translation steps in the TwoL system. for generating programs with mixed task parallel and data parallel executions on distributed memory machines (DMMs) based on the TwoL model [22, 23]. Previous papers on this programming model concentrate on the question whether for a specific application, the exploitation of task parallelism of a specific application leads to an efficient program for a given DMM and how design decisions like task scheduling and data distribution can be ....
T. Rauber and G. Runger. The Compiler TwoL for the Design of Parallel Implementations. In Proceedings of the 4th International Conference on Parallel Architectures and Compilation Techniques, Boston, IEEE, pages 282--301, 1996.
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