1 |
Evaluation de l'affectation des tâches sur une architecture à mémoire distribuée pour des modèles flot de données / Efficient evaluation of mappings of dataflow applications onto distributed memory architecturesLesparre, Youen 02 March 2017 (has links)
Avec l'augmentation de l'utilisation des smartphones, des objets connectés et des véhicules automatiques, le domaine des systèmes embarqués est devenu omniprésent dans notre environnement. Ces systèmes sont souvent contraints en terme de consommation et de taille. L'utilisation des processeurs many-cores dans des systèmes embarqués permet une conception rapide tout en respectant des contraintes temps-réels et en conservant une consommation énergétique basse.Exécuter une application sur un processeur many-core requiert un dispatching des tâches appelé problème de mapping et est connu comme étant NP-complet.Les contributions de cette thèse sont divisées en trois parties :Tout d'abord, nous étendons d'importantes propriétés dataflow au modèle Phased Computation Graph.Ensuite, nous présentons un générateur de graphe dataflow capable de générer des Synchonous Dataflow Graphs, Cyclo-Static Dataflow Graphs et Phased Computation Graphs vivant avec plus de 10000 tâches en moins de 30 secondes. Le générateur est comparé à SDF3 et PREESM.Enfin, la contribution majeure de cette thèse propose une nouvelle méthode d'évaluation d'un mapping en utilisant les modèles Synchonous Dataflow Graphe et Cyclo-Static Dataflow Graphe. La méthode évalue efficacement la mémoire consommée par les communications d'un dataflow mappé sur une architecture à mémoire distribuée. L'évaluation est déclinée en deux versions, la première garantit la vivacité alors que la seconde ajoute une contrainte de débit. La méthode d'évaluation est expérimentée avec des dataflow générés par Turbine et avec des applications réelles. / With the increasing use of smart-phones, connected objects or automated vehicles, embedded systems have become ubiquitous in our living environment. These systems are often highly constrained in terms of power consumption and size. They are more and more implemented with many-core processor array that allow, rapid design to meet stringent real-time constraints while operating at relatively low frequency, with reduced power consumption.Running an application on a processor array requires dispatching its tasks on the processors in order to meet capacity and performance constraints. This mapping problem is known to be NP-complete.The contributions of this thesis are threefold:First we extend important notions from the Cyclo-Static Dataflow Graph to the Phased Computation Graph model and two equivalent sufficient conditions of liveness.Second, we present a random dataflow graph generator able to generate Synchonous Dataflow Graphs, Cyclo-Static Dataflow Graphs and Phased Computation Graphs. The Generator, is able to generate live dataflow of up to 10,000 tasks in less than 30 seconds. It is compared with SDF3 and PREESM.Third and most important, we propose a new method of evaluation of a mapping using the Synchonous Dataflow Graph and the Cyclo-Static Dataflow Graph models. The method evaluates efficiently the memory footprint of the communications of a dataflow graph mapped on a distributed architecture. The evaluation is declined in two versions, the first guarantees a live mapping while the second accounts for a constraint on throughput.The evaluation method is experimented on dataflow graphs from Turbine and on real-life applications.
|
2 |
Mutli-objective trade-off exploration for Cyclo-Static and Synchronous Dataflow graphsSinha, Ashmita 30 October 2012 (has links)
Many digital signal processing and real-time streaming systems are modeled using dataflow graphs, such as Synchronous Dataflow (SDF) and Cyclo-static Dataflow (CSDF) graphs that allow static analysis and optimization techniques. However, mapping of such descriptions into tightly constrained real-time implementations requires optimization of resource sharing, buffering and scheduling across a multi-dimensional latency-throughput-area objective space. This requires techniques that can find the Pareto-optimal set of implementations for the designer to choose from. In this work, we address the problem of multi-objective mapping and scheduling of SDF and CSDF graphs onto heterogeneous multi-processor platforms. Building on previous work, this thesis extends existing two-stage hybrid heuristics that combine an evolutionary algorithm with an integer linear programming (ILP) model to jointly optimize throughput, area and latency for SDF graphs. The primary contributions of this work include: (1) extension of the ILP model to support CSDFGs with additional buffer size optimizations; (2) a further optimization in the ILP-based scheduling model to achieve a runtime speedup of almost a factor of 10 compared to the existing SDFG formulation; (3) a list scheduling heuristic that replaces the ILP model in the hybrid heuristic to generate Pareto-optimal solutions at significantly decreased runtime while maintaining near-optimality of the solutions within an acceptable gap of 10% when compared to its ILP counterparts. The list scheduling heuristic presented in this work is based on existing modulo scheduling approaches for software pipelining in the compiler domain, but has been extended by introducing a new concept of mobility-based rescheduling before resorting to backtracking. It has been proved in this work that if mobility-based rescheduling is performed, the number of required backtrackings and hence overall complexity and runtime is less. / text
|
Page generated in 0.0614 seconds