In this dissertation, we show that source-to-source optimization is an efficient method to generate a high performance program for irregular and heterogeneous code from a basic implementation. After describing the evolution of processor architectures, we provide two methods. The first one extract codelets from an irregular code, optimizing these codelets, and predicting the performance of the modified program. The other one limits the impact of alignment issues due to vectorization or bank conflicts. We also present two parallelization technics, one generating parallel codelets, the other scheduling a task graph on an heterogeneous system.Dans cette thèse, nous montrons que les optimisations source-à-source sont un moyen efficace pour génér...
The goal of this dissertation is to give programmers the ability to achieve high performance by focu...
Production compilers have achieved a high level of maturity in terms of generating efficient code. C...
Production compilers have achieved a high level of maturity in terms of generating efficient code. C...
In this dissertation, we show that source-to-source optimization is an efficient method to generate ...
In this dissertation, we show that source-to-source optimization is an efficient method to generate ...
Dans cette thèse, nous montrons que les optimisations source-à-source sont un moyen efficace pour gé...
The number of transistors as well as the frequency of processors have followed Moore's law for the p...
Although code optimizations are necessary to parallelize code, few guidelines exist for determining ...
The increasing complexity of processors has led to the development of a large number of code transfo...
The increasing complexity of processors has led to the development of a large number of code transfo...
The constant evolution of processors architectures, with superscalar, instruction-level parallelism,...
The constant evolution of processors architectures, with superscalar, instruction-level parallelism,...
Dans cette thèse, nous présentons nos contributions à APOLLO : un compilateur de parallélisation aut...
La compilation est une étape indispensable dans la création d'applications performantes.Cette étape ...
La complexité grandissante des architectures ne simplifie pas la tâche des compilateurs à générer du...
The goal of this dissertation is to give programmers the ability to achieve high performance by focu...
Production compilers have achieved a high level of maturity in terms of generating efficient code. C...
Production compilers have achieved a high level of maturity in terms of generating efficient code. C...
In this dissertation, we show that source-to-source optimization is an efficient method to generate ...
In this dissertation, we show that source-to-source optimization is an efficient method to generate ...
Dans cette thèse, nous montrons que les optimisations source-à-source sont un moyen efficace pour gé...
The number of transistors as well as the frequency of processors have followed Moore's law for the p...
Although code optimizations are necessary to parallelize code, few guidelines exist for determining ...
The increasing complexity of processors has led to the development of a large number of code transfo...
The increasing complexity of processors has led to the development of a large number of code transfo...
The constant evolution of processors architectures, with superscalar, instruction-level parallelism,...
The constant evolution of processors architectures, with superscalar, instruction-level parallelism,...
Dans cette thèse, nous présentons nos contributions à APOLLO : un compilateur de parallélisation aut...
La compilation est une étape indispensable dans la création d'applications performantes.Cette étape ...
La complexité grandissante des architectures ne simplifie pas la tâche des compilateurs à générer du...
The goal of this dissertation is to give programmers the ability to achieve high performance by focu...
Production compilers have achieved a high level of maturity in terms of generating efficient code. C...
Production compilers have achieved a high level of maturity in terms of generating efficient code. C...