Performance Effect of Localized Thread Schedules in Heterogeneous Multi-Core Processors
F. Sibai. Innovations in Information Technology, 2007. Innovations '07. 4th International Conference on, page 292-296. (2007)
DOI: 10.1109/IIT.2007.4430377
Abstract
We previously proposed a heterogeneous 16-core architecture and a nearest neighbor priority-based thread scheduling algorithm which values core affinity and permits inter-core thread migration in 1. In this paper, we run simulation experiments under localized core affinities limited to a very small number of specific cores within core classes. These localized schedules represent applications that are user- or system-confined to specific cores or their vicinity. They may also represent stressful scenarios with high contention to specific - and possibly scarce - cores with unique functionality or computational might. The results indicate that under these localized initial core affinity scenarios, non-migratory thread scheduling algorithms perform much worse than non-affinity algorithms while the algorithms which consider core affinity and allow inter-core thread migration are best.
%0 Conference Paper
%1 Sibai:2007
%A Sibai, F.N.
%B Innovations in Information Technology, 2007. Innovations '07. 4th International Conference on
%D 2007
%K heterogeneous multicore os-scheduling
%P 292-296
%R 10.1109/IIT.2007.4430377
%T Performance Effect of Localized Thread Schedules in Heterogeneous Multi-Core Processors
%U http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4430377
%X We previously proposed a heterogeneous 16-core architecture and a nearest neighbor priority-based thread scheduling algorithm which values core affinity and permits inter-core thread migration in 1. In this paper, we run simulation experiments under localized core affinities limited to a very small number of specific cores within core classes. These localized schedules represent applications that are user- or system-confined to specific cores or their vicinity. They may also represent stressful scenarios with high contention to specific - and possibly scarce - cores with unique functionality or computational might. The results indicate that under these localized initial core affinity scenarios, non-migratory thread scheduling algorithms perform much worse than non-affinity algorithms while the algorithms which consider core affinity and allow inter-core thread migration are best.
%@ 978-1-4244-1841-1
@inproceedings{Sibai:2007,
abstract = {We previously proposed a heterogeneous 16-core architecture and a nearest neighbor priority-based thread scheduling algorithm which values core affinity and permits inter-core thread migration in [1]. In this paper, we run simulation experiments under localized core affinities limited to a very small number of specific cores within core classes. These localized schedules represent applications that are user- or system-confined to specific cores or their vicinity. They may also represent stressful scenarios with high contention to specific - and possibly scarce - cores with unique functionality or computational might. The results indicate that under these localized initial core affinity scenarios, non-migratory thread scheduling algorithms perform much worse than non-affinity algorithms while the algorithms which consider core affinity and allow inter-core thread migration are best.},
added-at = {2008-05-06T17:30:41.000+0200},
author = {Sibai, F.N.},
biburl = {https://www.bibsonomy.org/bibtex/21d45c9111c2e79edd20401e5837d0c5a/ykwok},
booktitle = {Innovations in Information Technology, 2007. Innovations '07. 4th International Conference on},
doi = {10.1109/IIT.2007.4430377},
interhash = {662a116479de9302476ef4df4cc8ecd0},
intrahash = {1d45c9111c2e79edd20401e5837d0c5a},
isbn = {978-1-4244-1841-1},
keywords = {heterogeneous multicore os-scheduling},
pages = {292-296},
timestamp = {2008-05-06T17:30:41.000+0200},
title = {Performance Effect of Localized Thread Schedules in Heterogeneous Multi-Core Processors},
url = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4430377},
year = 2007
}