We are interested in the parallelization of a class of non-serial polyadic dynamic-programming problems in this thesis. These problems are characterized by a strong dependency between subproblems. To design efficient and portable parallel solutions to solve these problems, the CGM (coarse-grained multicomputer) model is the suitable choice because of its simplicity and its compatibility with most supercomputers.A CGM-based parallel algorithm is a succession of computation and communication rounds. The solutions proposed in the literature give the end-user the possibility of minimizing the number of communication rounds or balancing the load between processors because both objectives are conflicting. Moreover, their main drawback is to foste...
Technology trends are making the cost of data movement increasingly dominant, both in terms of energ...
This thesis deals with parallel graph partitioning and, more specifically, focuses on its applicatio...
Nous nous intéresserons au modèle systolique et au modèle parallèle à gros grains baptisé {Coarse Gr...
Several factors lead designers of parallel architectures to converge to coarse-grained multi-process...
Nous assistons cette décennie à une tendance (migration) du hardware parallèle vers les systèmes mul...
We have all the intuition that a work can be completed in much less time if it is distributed betwee...
Dynamic programming is a technique widely used to solve several combinatory optimization problems. A...
In this thesis we study the behavior of parallel applications represented by a precedence graph. The...
This thesis proposes several algorithms for graphs. The first chapter focuses on PRAM model. The sec...
Polyhedral techniques enable the application of analysis and code transformations on multi-dimension...
The aim of this thesis is the study of different methods to minimize the communication overhead due ...
Optimistic parallelization is a promising approach for the parallelization of irregular algorithms: ...
This thesis focus on the problem of scheduling the tasks of a parallel application taking into accou...
Scientific applications are commonly modeled as the processing of directed acyclicgraphs of tasks, a...
In this thesis, we study the state space of concurrent programs using the tools from analytic combin...
Technology trends are making the cost of data movement increasingly dominant, both in terms of energ...
This thesis deals with parallel graph partitioning and, more specifically, focuses on its applicatio...
Nous nous intéresserons au modèle systolique et au modèle parallèle à gros grains baptisé {Coarse Gr...
Several factors lead designers of parallel architectures to converge to coarse-grained multi-process...
Nous assistons cette décennie à une tendance (migration) du hardware parallèle vers les systèmes mul...
We have all the intuition that a work can be completed in much less time if it is distributed betwee...
Dynamic programming is a technique widely used to solve several combinatory optimization problems. A...
In this thesis we study the behavior of parallel applications represented by a precedence graph. The...
This thesis proposes several algorithms for graphs. The first chapter focuses on PRAM model. The sec...
Polyhedral techniques enable the application of analysis and code transformations on multi-dimension...
The aim of this thesis is the study of different methods to minimize the communication overhead due ...
Optimistic parallelization is a promising approach for the parallelization of irregular algorithms: ...
This thesis focus on the problem of scheduling the tasks of a parallel application taking into accou...
Scientific applications are commonly modeled as the processing of directed acyclicgraphs of tasks, a...
In this thesis, we study the state space of concurrent programs using the tools from analytic combin...
Technology trends are making the cost of data movement increasingly dominant, both in terms of energ...
This thesis deals with parallel graph partitioning and, more specifically, focuses on its applicatio...
Nous nous intéresserons au modèle systolique et au modèle parallèle à gros grains baptisé {Coarse Gr...