OpenACC Tutorial - Adding directives/fr: Difference between revisions

OpenACC Tutorial - Adding directives/fr (view source)

Revision as of 14:30, 11 May 2017

31 bytes added , 7 years ago

Created page with "Les utilisateurs d'[https://docs.computecanada.ca/wiki/OpenMP/fr OpenMP] retrouveront dans OpenACC le principe de directives. Il existe cependant d'importantes différences en..."

Diane27

rsnt_translations

56,430

edits

@@ Line 97: / Line 97: @@
 |title=Description ou prescription
 |content=
-Those who have used [[OpenMP]] before will be familiar with the directive based nature of OpenACC. There is however one major difference between OpenMP and OpenACC directives. OpenMP directives are by design ''prescriptive'' in nature. This means that the compiler is required to perform the requested parallelization, no matter whether this is good from a performance stand point or not. This yields very reproducible results from one compiler to the next. This also means that parallelization will be performed the same way, whatever the hardware the code runs on. However, not every architecture performs best with code written the same way. Sometimes, it may be beneficial to switch the order of loops for example. If one were to parallelize a code with OpenMP and wanted it to perform optimally on multiple different architectures, they would have to write different sets of directives for different architectures.
+Les utilisateurs d'[https://docs.computecanada.ca/wiki/OpenMP/fr OpenMP] retrouveront dans OpenACC le principe de directives. Il existe cependant d'importantes différences entre les directives OpenMP et OpenACC. Les directives OpenMP sont à la base ''prescriptives''. Ceci signifie que le compilateur est forcé d'accomplir la parallélisation, peu importe que l'effet détériore ou améliore la performance. Le résultat est prévisible pour tous les compilateurs. De plus, la parallélisation se fera de la même manière, peu importe le matériel utilisé pour exécuter le code. Par contre, le même code peut connaitre une moins bonne performance, dépendant de l'architecture. Il peut donc être préférable par exemple de changer l'ordre des boucles. Pour paralléliser du code avec OpenMP et obtenir une performance optimale dans différentes architectures, il faudrait avoir un ensemble différent de directives pour chaque architecture.
 In contrast, many of OpenACC's directives are ''descriptive'' in nature. This means that the compiler is free to compile the code whichever way it thinks is best for the target architecture. This may even imply that the code is not parallelized at all. The '''same code''', compiled to run on GPU, or on Xeon Phi, or on CPU, may therefore yield different binary code. This, of course, means that different compilers may yield different performance. It also means that new generations of compilers will do better than previous generations, especially with new hardware.