Use this link to cite:
http://hdl.handle.net/2183/32124 Optimización automática de PolyBench-Python
Loading...
Identifiers
Publication date
Authors
Piñeiro Bermúdez, Sergio
Advisors
Other responsabilities
Universidade da Coruña. Facultade de Informática
Journal Title
Bibliographic citation
Type of academic work
Academic degree
Abstract
[Resumen]: Este trabajo consiste en permitir la ejecución de los bancos de pruebas de PolyBench-Python con diferentes configuraciones que se obtienen de optimizaciones poliédricas conseguidas gracias al compilador poliédrico PoCC. De esta manera, se corrige y amplía la herramienta PolyPy que actúa como intermediador entre los benchmarks que forman PolyBench-Python y el compilador PoCC. Esto permitirá poder realizar un estudio empírico de las optimizaciones poliédricas sobre un lenguaje interpretado como Python. Por otra parte, también se desarrolla, dentro de PolyPy, una capa capaz de transformar el código de los benchmarks al formato NumPy. Esta configuración también es posible utilizarla en la ejecución de PolyBench-Python.
[Abstract]: This work consists of allowing the execution of the PolyBench-Python benchmarks with different configurations that are obtained from polyhedral optimizations that are achieved thanks to the PoCC polyhedral compiler. In this way, the PolyPy tool that acts as an intermediary between the benchmarks that make up PolyBench-Python and the PoCC compiler is corrected and extended. This will allow us to carry out an empirical study of polyhedral optimizations on an interpreted language such as Python. On the other hand, a layer capable of transforming the code of the benchmarks to the NumPy format is also developed within PolyPy. This configuration can also be used in the execution of PolyBench-Python.
[Abstract]: This work consists of allowing the execution of the PolyBench-Python benchmarks with different configurations that are obtained from polyhedral optimizations that are achieved thanks to the PoCC polyhedral compiler. In this way, the PolyPy tool that acts as an intermediary between the benchmarks that make up PolyBench-Python and the PoCC compiler is corrected and extended. This will allow us to carry out an empirical study of polyhedral optimizations on an interpreted language such as Python. On the other hand, a layer capable of transforming the code of the benchmarks to the NumPy format is also developed within PolyPy. This configuration can also be used in the execution of PolyBench-Python.
Description
Editor version
Rights
Atribución-NoComercial-SinDerivadas 3.0 España







