J
Jean-Louis Roch
Researcher at University of Grenoble
Publications - 78
Citations - 689
Jean-Louis Roch is an academic researcher from University of Grenoble. The author has contributed to research in topics: Parallel algorithm & Matrix multiplication. The author has an hindex of 15, co-authored 78 publications receiving 673 citations. Previous affiliations of Jean-Louis Roch include Centre national de la recherche scientifique & French Institute for Research in Computer Science and Automation.
Papers
More filters
Proceedings ArticleDOI
Athapascan-1: On-line building data flow graph in a parallel language
TL;DR: This work introduces a language named Athapascan-1 that allows to build a graph of dependencies from a strong typing of shared memory accesses and shows that such a data dependency graph can be computed on-line on a distributed architecture.
Book ChapterDOI
A Tighter Analysis of Work Stealing
TL;DR: A general methodology for computing the expected makespan based on the analysis of an adequate potential function which represents the load unbalance between the local lists is presented and a bound on the deviation from the mean is derived.
Book ChapterDOI
Deque-Free Work-Optimal Parallel STL Algorithms
TL;DR: This paper presents provable work-optimal parallelizations of STL (Standard Template Library) algorithms based on the work-stealing technique with both theoretical and experimental results bounding the work/running time.
Book ChapterDOI
Regular versus Irregular Problems and Algorithms
TL;DR: A classification based on regularity criteria i.e. measures of how much an algorithm is regular (or irregular) is proposed, which expresses more the quality of the schedules that can be found as opposed to the way the schedules are obtained.
Book ChapterDOI
A checkpoint/recovery model for heterogeneous dataflow computations using work-stealing
TL;DR: A new checkpoint/recovery method for dataflow computations using work-stealing in heterogeneous environments as found in grid or cluster computing is presented, showing that both methods have very small overhead and that trade-offs between checkpointing and recovery cost can be controlled.