Parallelizing Legacy Fortran Programs Using Rewriting Rules Technique and Algebraic Program Models
| Autor: | Anatoliy Doroshenko, Kostiantyn Zhereb |
|---|---|
| Rok vydání: | 2013 |
| Předmět: |
Multi-core processor
Source code Computer science Fortran Programming language media_common.quotation_subject Process (computing) Parallel computing computer.software_genre Automatic parallelization Transformation (function) Code (cryptography) Rewriting computer computer.programming_language media_common |
| Zdroj: | ICT in Education, Research, and Industrial Applications ISBN: 9783642357367 |
| DOI: | 10.1007/978-3-642-35737-4_3 |
| Popis: | We present ongoing research in the area of transforming existing sequential Fortran programs into their parallel equivalents. We propose a semi-automated parallelization approach that uses rewriting rules technique to automate certain steps of the transformation process. A sequential source code is transformed into a parallel code for shared-memory systems, such as multicore processors. Parallelizing and optimizing transformations are formally described as rewriting rules which allows their automated application across the whole source code, and also facilitates their implementation and reuse. Using high-level algebraic models allows to describe program transformations in a more concise and stepwise manner. Performance measurements demonstrate the high efficiency of the obtained parallel programs, compared to the initial sequential programs and also to automated parallelization tools. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|