Exploiting the full computational power of current hierarchical multiprocessor machines requires a very careful distribution
of threads and data among the underlying non-uniform architecture so as to avoid memory access penalties. Directive-basedprogramming languages such as OpenMPprovide programmers with an easy way to structure the parallelism of their applicationand to transmit this information to the runtime system.
%0 Journal Article
%1 françois2009dynamic
%A Broquedis, François
%A Furmento, Nathalie
%A Goglin, Brice
%A Namyst, Raymond
%A Wacrenier, Pierre-André
%D 2009
%J Evolving OpenMP in an Age of Extreme Parallelism
%K Hierarchical Memory NUMA OpenMP Optimization Runtime Scheduling Thread
%P 79--92
%T Dynamic Task and Data Placement over NUMA Architectures: An OpenMP Runtime Perspective
%U http://dx.doi.org/10.1007/978-3-642-02303-3_7
%X Exploiting the full computational power of current hierarchical multiprocessor machines requires a very careful distribution
of threads and data among the underlying non-uniform architecture so as to avoid memory access penalties. Directive-basedprogramming languages such as OpenMPprovide programmers with an easy way to structure the parallelism of their applicationand to transmit this information to the runtime system.
@article{françois2009dynamic,
abstract = {Exploiting the full computational power of current hierarchical multiprocessor machines requires a very careful distribution
of threads and data among the underlying non-uniform architecture so as to avoid memory access penalties. Directive-basedprogramming languages such as OpenMPprovide programmers with an easy way to structure the parallelism of their applicationand to transmit this information to the runtime system.},
added-at = {2009-11-12T16:57:11.000+0100},
author = {Broquedis, François and Furmento, Nathalie and Goglin, Brice and Namyst, Raymond and Wacrenier, Pierre-André},
biburl = {https://www.bibsonomy.org/bibtex/2fd0deda4c4fffcb52df86235b658aca3/gron},
description = {SpringerLink - Book Chapter},
interhash = {bb2bed630731995727e1f51c54f28d9a},
intrahash = {fd0deda4c4fffcb52df86235b658aca3},
journal = {Evolving OpenMP in an Age of Extreme Parallelism},
keywords = {Hierarchical Memory NUMA OpenMP Optimization Runtime Scheduling Thread},
pages = {79--92},
timestamp = {2009-11-12T16:57:12.000+0100},
title = {Dynamic Task and Data Placement over NUMA Architectures: An OpenMP Runtime Perspective},
url = {http://dx.doi.org/10.1007/978-3-642-02303-3_7},
year = 2009
}