Memory-aware tree partitioning on homogeneous platforms

نویسندگان

  • Changjiang Gou
  • Anne Benoit
  • Loris Marchal
چکیده

Scientific applications are commonly modeled as the processing of directed acyclic graphs of tasks, and for some of them, the graph takes the special form of a rooted tree. This tree expresses both the computational dependencies between tasks and their storage requirements. The problem of scheduling/traversing such a tree on a single processor to minimize its memory footprint has already been widely studied. Hence, we move to parallel processing and study how to partition the tree for a homogeneous multiprocessor platform, where each processor is equipped with its own memory. We formally state the problem of partitioning the tree into subtrees such that each subtree can be processed on a single processor and the total resulting processing time is minimized. We prove that the problem is NP-complete, and we design polynomial-time heuristics to address it. An extensive set of simulations demonstrates the usefulness of these heuristics. Key-words: Scheduling, graph partitioning, memory-aware, makespan minimization Partitionnement d’arbres de tâches orienté mémoire pour les plates-formes homogènes Résumé : Les applications scientifiques sont couramment modélisées par des graphes de tâches. Pour certaines d’entre elles, le graphe prend la forme particulière d’un arbre enraciné. Cet arbre détermine à la fois les dépendance entre tâches de calcul et les besoins en stockage. Le problème d’ordonnancer (ou parcourir) un tel arbre sur un seul processeur pour réduire son empreinte mémoire a déjà largement été étudié. Dans ce rapport, nous considérons le traitement parallèle d’un tel arbre et étudions comment le partitionner pour une plate-forme de calcul formée de processeurs homogènes disposant chacun de sa propre mémoire. Nous formalisons le problème du partitionnement de l’arbre en sous-arbres de telle sorte que chaque sous-arbre puisse être traité sur un seul processeur et que le temps de calcul total soit minimal. Nous montrons que ce problème est NP-complet et proposons des heuristiques polynomiales. Un ensemble exhaustif de simulations permet de montrer l’utilité de ces heuristiques. Mots-clés : Ordonnancement, partitionnement de graphe, algorithmes orientés mémoire Memory-aware tree partitioning on homogeneous platfroms 3

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Performance and Energy Aware Workload Partitioning on Heterogeneous Platforms

Heterogeneous platforms which employ a mix of CPUs and accelerators such as GPUs have been widely used in the high-performance computing area [1]. Such heterogeneous platforms have the potential to offer higher performance at lower energy cost than homogeneous platforms. However, it is rather challenging to actually achieve the high performance and energy efficiency promised by heterogeneous pl...

متن کامل

Parallel Processing Algorithms for Content-based Retrieval from a Multimedia Database

The paper proposes a data partitioning algorithm for content-based retrieval from a multimedia database using a homogeneous cluster system. The multimedia data are represented in a unified model formed into a hierarchical multidimensional k-tree structure. The parallel retrieval algorithms run on a cluster of homogeneous processing units using a dedicated type of shared memory. We use system ch...

متن کامل

Energy-Aware Task Partitioning on Heterogeneous Multiprocessor Platforms

Efficient task partitioning plays a crucial role in achieving high performance at multiprocessor platforms. This paper addresses the problem of energy-aware static partitioning of periodic realtime tasks on heterogeneous multiprocessor platforms. A Particle Swarm Optimization variant based on Min-min technique for task partitioning is proposed. The proposed approach aims to minimize the overall...

متن کامل

Parallel Tree Building on a Range of Shared Address Space Multiprocessors: Algorithms and Application Performance

Irregular, particle-based applications that use trees, for example hierarchical N-body applications, are important consumers of multiprocessor cycles, and are argued to benefit greatly in programming ease from a coherent shared address space programming model. As more and more supercomputing platforms that can support different programming models become available to users, from tightly-coupled ...

متن کامل

FAST: A Generic Framework for Flash-Aware Spatial Trees

Spatial tree index structures are crucial components in spatial data management systems, designed with the implicit assumption that the underlying external memory storage is the conventional magnetic hard disk drives. This assumption is going to be invalid soon, as flash memory storage is increasingly adopted as the main storage media in mobile devices, digital cameras, embedded sensors, and no...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2017