Building and Scheduling Coarse Grain Task Graphs

نویسندگان

  • M. Cosnard
  • E. Jeannot
چکیده

It has been shown recently that it is possible to eeciently schedule direct acyclic task graphs. A tool like Pyrros 14] is able to generate a parallel program, if the user speciies precedence constraints between the tasks. This approach has two major drawbacks: it is hard for the user to perform a good analysis of the source code and scheduling a DAG is costly in terms of memory and time. In 3], a new model of parallel computation, the parameterized task graph is introduced. The Parameterized Task Graph (PTG) is a compact, problem size independent representation of some frequently used direct acyclic task graphs. In 4, 10], techniques and a tool (PlusPyr), are designed to generate such a representation from an annotated sequential program. In this report, we review the PTG and associated techniques, we give a general framework and an algorithm to dynamically schedule the parameterized graph and generate the associated code. On a montr e r ecemment qu'il est possible d'ordonnancer eecacement un graphe de t^ aches acyclique. Un outil comme Pyrros 14] est capable de g en erer un programme parall ele si l'utilisateur sp eciie les contraintes de pr ec edence entre les t^ aches. Cette approche a deux d efauts majeurs : il est diicile pour l'utilisateur de faire une bonne analyse du code source et ordonnancer un graphe acyclique est co^ uteux en termes de temps et de m emoire. Dans 3], un nouveau mod ele de calcul parall ele, le graphe de t^ aches param etr e a et e introduit. Le GTP est une repr esentation compacte, ind ependante de la taille du probl eme, de quelques graphes de t^ aches acycliques orient es fr equemment utilis es. Dans 4, 10], des techniques et un outil (PlusPyr) sont construit pour g en erer une telle repr esentation a partir d'un programme s equentiel annot e. Dans ce rapport, nous revoyons le GTP et les techniques associ es, nous donnons une structure g en erale et un algorithme pour ordonnancer dynamiquement le graphe param etr e et g en erer le code associ e.

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

Low Memory Cost Dynamic Scheduling of Large Coarse Grain Task Graphs

Scheduling large task graphs is an important issue in parallel computing since it allows the treatment of big size problems. In this paper we tackle the following problem: how to schedule a task graph, when it is too large to fit into memory? Our answer features the parameterized task graph (PTG), which is a symbolic representation of the task graph. We propose a dynamic scheduling algorithmwhi...

متن کامل

Symbolic Partitioning and Scheduling of Parameterized Task Graphs

The DAG-based task graph model has been found effective in scheduling for performance prediction and optimization of parallel applications. However the scheduling complexity and solution normally depend on the problem size. In this paper, we propose a symbolic scheduling scheme for a parameterized task graph which models coarse-grain DAG parallelism independent of the problem size. The algorith...

متن کامل

SLC: Symbolic Scheduling for Executing Parameterized Task Graphs on Multiprocessors

Task graph scheduling has been found effective in performance prediction and optimization of parallel applications. A number of static scheduling algorithms have been proposed for task graph execution on distributed memory machines. Such an approach cannot be adapted to changes in values of program parameters and the number of processors and also it cannot handle large task graphs. In this pape...

متن کامل

CPR: Mixed Task and Data Parallel Scheduling for Distributed Systems

It is well-known that mixing task and data parallelism to solve large computational applications often yields better speedups compared to either applying pure task parallelism or pure data parallelism. Typically, the applications are modeled in terms of a dependence graph of coarse-grain data-parallel tasks, called a data-parallel task graph. In this paper we present a new compile-time heuristi...

متن کامل

LLB: A Fast and Effective Scheduling Algorithm for Distributed-Memory Systems

This paper presents a new algorithm called List-based Load Balancing (LLB) for compile-time task scheduling on distributed-memory machines. LLB is intended as a cluster-mapping and task-ordering step in the multi-step class of scheduling algorithms. Unlike current multistep approaches, LLB integrates cluster-mapping and taskordering in a single step. The benefits of this integration are twofold...

متن کامل

An Efficient Genetic Algorithm for Task Scheduling on Heterogeneous Computing Systems Based on TRIZ

An efficient assignment and scheduling of tasks is one of the key elements in effective utilization of heterogeneous multiprocessor systems. The task scheduling problem has been proven to be NP-hard is the reason why we used meta-heuristic methods for finding a suboptimal schedule. In this paper we proposed a new approach using TRIZ (specially 40 inventive principles). The basic idea of thi...

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 1997