The classical redistribution problem aims at optimally scheduling communications when moving from an initial data distribution \Dini to a target distribution \Dtar where each processor $P_{i}$ will host a subset $P(i)$ of data items. However, modern computing platforms are equipped with a powerful interconnection switch, and the cost of a given communication is (almost) independent of the location of its sender and receiver. This leads to generalizing the redistribution problem as follows: find the optimal permutation $\sigma$ of processors such that $P_{i}$ will host the set $P(\sigma(i))$, and for which the cost of the redistribution is minimal. This report studies the complexity of this generalized problem. We provide optimal algorithms ...
This paper presents communication-efficient algorithms for the external data redistribution problem....
This report discusses a new algorithm for makespan minimization on situations where the workload can...
Many complicated parallel computing applications are composed of several stages. As the program proc...
The classical redistribution problem aims at optimally scheduling communications when moving from an...
International audienceThe classical redistribution problem aims at optimally scheduling communicatio...
The classical redistribution problem aims at optimally scheduling communications when moving from an...
We consider the problem of redistributing data on homogeneous and heterogeneous ring of processors. ...
Nous considérons ici le problème où deux programmes différents situés sur deux grappes d'ordinateurs...
In this thesis, we study iterative algorithms onto heterogeneous platforms. These iterative algorith...
Abstract-Decentralized algorithms are examined for opti-mally distributing a divisible resource in a...
Implementing linear algebra kernels on distributed memory parallel computers raises the problem of d...
Dans cette thèse, nous nous sommes intéressée à la mise en oeuvre d'algorithmes itératifs sur des gr...
A major issue of the operation of distributed systems is the problem of allocating a number of proce...
Reduction is a core operation in parallel computing. Optimizing its cost has a high potential impact...
La connectivité totale offerte par la communication sans fil pose un grand nombre d'avantages et de ...
This paper presents communication-efficient algorithms for the external data redistribution problem....
This report discusses a new algorithm for makespan minimization on situations where the workload can...
Many complicated parallel computing applications are composed of several stages. As the program proc...
The classical redistribution problem aims at optimally scheduling communications when moving from an...
International audienceThe classical redistribution problem aims at optimally scheduling communicatio...
The classical redistribution problem aims at optimally scheduling communications when moving from an...
We consider the problem of redistributing data on homogeneous and heterogeneous ring of processors. ...
Nous considérons ici le problème où deux programmes différents situés sur deux grappes d'ordinateurs...
In this thesis, we study iterative algorithms onto heterogeneous platforms. These iterative algorith...
Abstract-Decentralized algorithms are examined for opti-mally distributing a divisible resource in a...
Implementing linear algebra kernels on distributed memory parallel computers raises the problem of d...
Dans cette thèse, nous nous sommes intéressée à la mise en oeuvre d'algorithmes itératifs sur des gr...
A major issue of the operation of distributed systems is the problem of allocating a number of proce...
Reduction is a core operation in parallel computing. Optimizing its cost has a high potential impact...
La connectivité totale offerte par la communication sans fil pose un grand nombre d'avantages et de ...
This paper presents communication-efficient algorithms for the external data redistribution problem....
This report discusses a new algorithm for makespan minimization on situations where the workload can...
Many complicated parallel computing applications are composed of several stages. As the program proc...