journée du groupe de travail LaMHA - Automne 2018

Langages et Modèles de Haut-niveau pour la programmation parallèle, distribuée, de grilles de calcul et Applications
Groupe de travail du GdR GPL
Jeudi 13 décembre 2018
Sorbonne Université - Campus Pierre et Marie Curie
Rotonde 25, 4ème étage, couloir 24-25, porte 405
4 place Jussieu - 75005 Paris
Accès au campus Jussieu

Programme de la journée

10h00 Accueil
10h30 Emmanuelle Saillard, oratrice invitée.
Inria Bordeaux, STORM
icone pdf Analyse statique/dynamique pour la vérification des applications parallèles.
11h Julien Tesson
Nomadic Labs
icone pdf Interface web pour l'apprentissage de la programmation parallèle avec BSML
11h30 Lionel Lacassagne, orateur invité.
LIP6, ALSOC
Parallélisation d'Union Find sur CPU multicoeurs, sur GPU et sur CPU multi-coeurs SIMD: application a l'étiquetage en composantes connexes.
12h00 Déjeuner
14h Ludovic Henrio
LIP/Cash
presentation Ludovic HENRIO format powerpoint Ensuring Memory Consistency in Heterogeneous Systems Based on Access Mode Declarations.
14h30 Wijnand Suijlen
Huawei Technologies France / Paris Research Center
Parallel algorithms

In my past time I like to watch old movies, not because they are old per se, but because they must be really good if after all that time, people still remember them. Similarly, I like old parallel algorithms. In this talk I like to talk about a combination of two classics: Bruck's all-to-all and Valiant's two-phase randomized routing. Both describe fast ways of communicating between processes in a parallel computer. By combining the two you get a new communication algorithm, which effectively transforms a high-throughput network into a low-latency network. Although Bruck's and Valiant's algorithms are true classics, I will present them both and, of course, also their combination: Randomized Bruck. Cost analysis is included, but also demonstrated with some benchmark results on a small HPC cluster.

15h Pause café
15h30 Victor Allombert
Nomadic Labs
icone pdf Sémantique opérationnelle avec modèle de coût pour Multi-ML

BSML and Multi-ML are functional parallel programming languages “à la ML” based of the respectively the bsp and Multi-BSP bridging models. Multi-BSP extends bsp to take into account hierarchical architectures. For both models, it is possible to predict the performances of algorithms thanks to embedded cost models. To do so, we propose formal a operational semantics with cost annotations for the two aforementioned languages. This work has been done in a incremental manner. First we recall the cost semantics of core-ml language. Then, we adapt it to bsml and we adapt it to Multi-ML. It is then possible to evaluate the cost of a program following the annotated semantics. Finally, we compare the theoretical approach with the current implementation on a code example.

16h00 Sylvain Jubertie, LIFO PAMDA Constructions d'abstractions en C++ basées sur des templates et transformations par les compilateurs.
16h30-17h30 Pot et discussions

Inscription

Gratuite mais nécessaire, l'inscription doit être faite avant le 7 décembre par email aux organisateurs.

Accès

Contact

Emmanuel Chailloux
Sylvain Jubertie