8645 articles  [version française]
For submit on Hal-Inria: http://hal.inria.fr

Contact : archive-ouverte@inria.fr

Last report deposit of Inria teams

Publications HAL de la collection INRIA-RRRT

2014

Titre
On the proper orientation number of bipartite graphs
Auteurs
Julio Araujo ; Nathann Cohen; Susanna F. De Rezende; Frederic Havet; Phablo Moura
Détail
[Research Report], 2014, pp. 23. RR-8492
Début du résumé
An {\it orientation} of a graph~$G$ is a digraph~$D$ obtained from~$G$ by replacing each edge by exactly one of the two possible arcs with the same endvertices. For each~$v \in V(G)$, the \emph{indegree} of~$v$ in~$D$, denoted by~$d^-_D(v)$, is the number of arcs with head~$v$ in~$D$. An orientation~$D$ of~$G$ is \emph{proper} if~$d^-_D(u)\neq d^-_D(v)$, for all~$uv\in E(G)$. The \emph{proper orientation number} of a graph~$G$, denoted by~$\po(G)$, is the minimum of the maximum indegree over all its proper orientations. In this paper, we prove that~$\po(G) \leq \left\lfloor \left(\Delta(G) + \sqrt{\Delta(G)}\right)/2 \right\rfloor+ 1$ if~$G$ is a bipartite graph, and~$\po(G)\leq 4$ if~$G$ is a tree. .....
Accès au texte intégral et bibtex
Titre
Locality optimization on a NUMA architecture for hybrid LU factorization
Auteurs
Adrien Rémy; Marc Baboulin ; Masha Sosonkina; Brigitte Rozoy
Détail
[Research Report], 2014. RR-8497
Début du résumé
We study the impact of non-uniform memory accesses (NUMA) on the solution of dense general linear systems using an LU factorization algorithm. In particular we illustrate how an appropriate placement of the threads and memory on a NUMA architecture can improve the performance of the panel factorization and consequently accelerate the global LU factorization. We apply these placement strategies and present performance results for a hybrid multicore/GPU LU algorithm as it is implemented in the public domain library MAGMA. .....
Accès au texte intégral et bibtex
Titre
Continuously Measuring Critical Section Pressure with the Free Lunch Profiler
Auteurs
Florian David; Gaël Thomas; Julia Lawall; Gilles Muller
Détail
Mar. 2014, RR-8486
Début du résumé
Today, Java is regularly used to implement large multi-threaded server-class ap- plications, whose scalability could be hampered by the execution of critical sections. Profiling is needed to identify critical sections that are problematic. However, profiling such applications is challenging, due to their long running times and the range of possible runtime conditions. We propose Free Lunch, a new profiler designed to identify locks and critical sections that hamper scalability. Free Lunch is designed around a new metric, critical section pressure, and can be used in-vivo, while the application is run by end-users. Using Free Lunch, we have identified a contention .....
Accès au texte intégral et bibtex
Titre
A Unifying View of Loosely Time-Triggered Architectures
Auteurs
Guillaume Baudart ; Albert Benveniste; Anne Bouillard; Paul Caspi
Détail
[Research Report], 2014, pp. 14. RR-8494
Début du résumé
Cyber-Physical Systems require distributed architectures to support safety critical real-time control. Hermann Kopetz' Time-Triggered Architecture (TTA) has been proposed as both an architecture and a comprehensive paradigm for systems architecture, for such systems. TTA offers the programmer a logical discrete time compliant with synchronous programming, together with timing bounds. A clock synchronization protocol is required, unless the local clocks used themselves provide the recquired accuracy. To relax the strict requirements on synchronization imposed by TTA, Loosely Time-Triggered Architectures (LTTA) have been proposed. In LTTA, computation and communication units are all triggered by autonomous, unsynchronized, clocks. Communication media act as shared .....
Accès au texte intégral et bibtex
Titre
Asymptotic description of neural networks with correlated synaptic weights
Auteurs
Olivier Faugeras ; James Maclaurin
Détail
[Research Report], 2014, pp. 47. RR-8495
Début du résumé
We study the asymptotic law of a network of interacting neurons when the number of neurons becomes infinite. Given a completely connected network of neurons in which the synaptic weights are Gaussian {\emph correlated} random variables, we describe the asymptotic law of the network when the number of neurons goes to infinity. All previous works assumed that the weights were i.i.d. random variables, thereby making the analysis much simpler. This hypothesis is not realistic from the biological viewpoint. In order to cope with this extra complexity we introduce the process-level empirical measure of the trajectories of the solutions to the .....
Accès au texte intégral et bibtex