The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Un roman Partitionnement à plusieurs niveaux (MLP) prenant en compte contraintes du monde réel pour le partitionnement matériel-logiciel dans Systèmes multiprocesseurs embarqués distribués (DEMS) est proposé. Cet algorithme MLP utilise une métrique de gradient basée sur le coût et les performances du matériel et du logiciel comme métrique de base pour la sélection des partitions optimales et se compose de trois partitions imbriquées. niveaux. Le niveau le plus interne est une simple recherche binaire qui permet des évaluations rapides d'un grand nombre de partitions possibles. Le niveau intermédiaire parcourt différentes allocations possibles de processeurs (qui exécutent le logiciel) aux sous-systèmes. Le niveau le plus externe itère sur le nombre de processeurs et la fourchette de coûts du matériel. Des heuristiques sont appliquées à chaque niveau pour éviter la recherche exhaustive coûteuse. L'application du MLP comme objectif récemment Conception de code de système embarqué distribué (DESC) montre sa faisabilité. Les comparaisons entre des exemples réels partitionnés à l'aide de MLP et à l'aide d'autres techniques existantes démontrent les atouts contrastés de MLP. Le partage, le clustering et le modèle de système hiérarchique sont quelques fonctionnalités importantes de MLP, qui contribuent à produire des résultats de partitionnement plus optimaux.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copier
Trong-Yen LEE, Pao-Ann HSIUNG, Sao-Jie CHEN, "Hardware-Software Multi-Level Partitioning for Distributed Embedded Multiprocessor Systems" in IEICE TRANSACTIONS on Fundamentals,
vol. E84-A, no. 2, pp. 614-626, February 2001, doi: .
Abstract: A novel Multi-Level Partitioning (MLP) technique taking into account real-world constraints for hardware-software partitioning in Distributed Embedded Multiprocessor Systems (DEMS) is proposed. This MLP algorithm uses a gradient metric based on hardware-software cost and performance as the core metric for selection of optimal partitions and consists of three nested levels. The innermost level is a simple binary search that allows quick evaluations of a large number of possible partitions. The middle level iterates over different possible allocations of processors (that execute software) to subsystems. The outermost level iterates over the number of processors and the hardware cost range. Heuristics are applied to each level to avoid the expensive exhaustive search. The application of MLP as a recently purposed Distributed Embedded System Codesign (DESC) methodology shows its feasibility. Comparisons between real-world examples partitioned using MLP and using other existing techniques demonstrate contrasting strengths of MLP. Sharing, clustering, and hierarchical system model are some important features of MLP, which contribute towards producing more optimal partition results.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/e84-a_2_614/_p
Copier
@ARTICLE{e84-a_2_614,
author={Trong-Yen LEE, Pao-Ann HSIUNG, Sao-Jie CHEN, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Hardware-Software Multi-Level Partitioning for Distributed Embedded Multiprocessor Systems},
year={2001},
volume={E84-A},
number={2},
pages={614-626},
abstract={A novel Multi-Level Partitioning (MLP) technique taking into account real-world constraints for hardware-software partitioning in Distributed Embedded Multiprocessor Systems (DEMS) is proposed. This MLP algorithm uses a gradient metric based on hardware-software cost and performance as the core metric for selection of optimal partitions and consists of three nested levels. The innermost level is a simple binary search that allows quick evaluations of a large number of possible partitions. The middle level iterates over different possible allocations of processors (that execute software) to subsystems. The outermost level iterates over the number of processors and the hardware cost range. Heuristics are applied to each level to avoid the expensive exhaustive search. The application of MLP as a recently purposed Distributed Embedded System Codesign (DESC) methodology shows its feasibility. Comparisons between real-world examples partitioned using MLP and using other existing techniques demonstrate contrasting strengths of MLP. Sharing, clustering, and hierarchical system model are some important features of MLP, which contribute towards producing more optimal partition results.},
keywords={},
doi={},
ISSN={},
month={February},}
Copier
TY - JOUR
TI - Hardware-Software Multi-Level Partitioning for Distributed Embedded Multiprocessor Systems
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 614
EP - 626
AU - Trong-Yen LEE
AU - Pao-Ann HSIUNG
AU - Sao-Jie CHEN
PY - 2001
DO -
JO - IEICE TRANSACTIONS on Fundamentals
SN -
VL - E84-A
IS - 2
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - February 2001
AB - A novel Multi-Level Partitioning (MLP) technique taking into account real-world constraints for hardware-software partitioning in Distributed Embedded Multiprocessor Systems (DEMS) is proposed. This MLP algorithm uses a gradient metric based on hardware-software cost and performance as the core metric for selection of optimal partitions and consists of three nested levels. The innermost level is a simple binary search that allows quick evaluations of a large number of possible partitions. The middle level iterates over different possible allocations of processors (that execute software) to subsystems. The outermost level iterates over the number of processors and the hardware cost range. Heuristics are applied to each level to avoid the expensive exhaustive search. The application of MLP as a recently purposed Distributed Embedded System Codesign (DESC) methodology shows its feasibility. Comparisons between real-world examples partitioned using MLP and using other existing techniques demonstrate contrasting strengths of MLP. Sharing, clustering, and hierarchical system model are some important features of MLP, which contribute towards producing more optimal partition results.
ER -