Julien Herrmann

32
Documents
Affiliations actuelles
  • Centre National de la Recherche Scientifique (CNRS)
  • Algorithmes Parallèles et Optimisation (IRIT-APO)
  • Institut de recherche en informatique de Toulouse (IRIT)

Publications

Publications

Deposit thumbnail

Optimal checkpointing for heterogeneous chains: how to train deep neural networks with limited memory Optimal Re-Materialization Strategies for Heterogeneous Chains: How to Train Deep Neural Networks with Limited Memory

Olivier Beaumont , Lionel Eyraud-Dubois , Julien Herrmann , Alexis Joly , Alena Shilova

ACM Transactions on Mathematical Software, 2024, 50 (2), pp.1-38. ⟨10.1145/3648633⟩

Article dans une revue hal-02352969v2
Deposit thumbnail

checkpoint_schedules: schedules for incremental checkpointing of adjoint simulations

Daiane I Dolci , James R Maddison , David A Ham , Guillaume Pallez , Julien Herrmann

Journal of Open Source Software, 2024, 9, pp.1-4. ⟨10.21105/joss.06148⟩

Article dans une revue hal-04854312v1
Deposit thumbnail

Task-based parallel programming for scalable matrix product algorithms

Emmanuel Agullo , Alfredo Buttari , Abdou Guermouche , Julien Herrmann , Antoine Jego

ACM Transactions on Mathematical Software, 2023, 49 (2), pp.Pages 1 - 23. ⟨10.1145/3583560⟩

Article dans une revue hal-03936659v2
Deposit thumbnail

H-Revolve: A Framework for Adjoint Computation on Synchronous Hierarchical Platforms

Julien Herrmann , Guillaume Pallez

ACM Transactions on Mathematical Software, 2020, ⟨10.1145/3378672⟩

Article dans une revue hal-02080706v2
Deposit thumbnail

Multilevel Algorithms for Acyclic Partitioning of Directed Acyclic Graphs

Julien Herrmann , Yusuf M. Özkaya , Bora Uçar , Kamer Kaya , Ümit V. Çatalyürek

SIAM Journal on Scientific Computing, 2019, 41 (4), pp.A2117-A2145. ⟨10.1137/18M1176865⟩

Article dans une revue hal-02306566v1
Deposit thumbnail

Optimal Memory-aware Backpropagation of Deep Join Networks

Olivier Beaumont , Julien Herrmann , Guillaume Pallez , Alena Shilova

Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, In press

Article dans une revue hal-02401105v1
Deposit thumbnail

Computing the expected makespan of task graphs in the presence of silent errors

Henri Casanova , Julien Herrmann , Yves Robert

Parallel Computing, 2018, 75, pp.41-60. ⟨10.1016/j.parco.2018.03.004⟩

Article dans une revue hal-01968433v1
Deposit thumbnail

Periodicity in optimal hierarchical checkpointing schemes for adjoint computations

Guillaume Aupy , Julien Herrmann

Optimization Methods and Software, 2017, 32 (3), pp.594-624. ⟨10.1080/10556788.2016.1230612⟩

Article dans une revue hal-01654632v1
Deposit thumbnail

Assessing the cost of redistribution followed by a computational kernel: Complexity and performance results

Julien Herrmann , George Bosilca , Thomas Hérault , Loris Marchal , Yves Robert et al.

Parallel Computing, 2016, 52, pp.20. ⟨10.1016/j.parco.2015.09.005⟩

Article dans une revue hal-01254167v1
Deposit thumbnail

Optimal Multistage Algorithm for Adjoint Computation

Guillaume Aupy , Julien Herrmann , Paul Hovland , Yves Robert

SIAM Journal on Scientific Computing, 2016, 38 (3), pp.C232-C255. ⟨10.1137/15M1019222⟩

Article dans une revue hal-01354902v1

Memory-aware tree traversals with pre-assigned tasks

Julien Herrmann , Loris Marchal , Yves Robert

Journal of Parallel and Distributed Computing, 2015, 75, pp.53--56

Article dans une revue hal-01026405v1

Mixing LU and QR factorization algorithms to design high-performance dense linear algebra solvers

Mathieu Faverge , Julien Herrmann , Julien Langou , Bradley Lowery , Yves Robert et al.

Journal of Parallel and Distributed Computing, 2015, IPDPS 2014 Selected Papers on Numerical and Combinatorial Algorithms, 85, pp.15. ⟨10.1016/j.jpdc.2015.06.007⟩

Article dans une revue hal-01107457v1

Accelerating linear system solutions using randomization technique

Marc Baboulin , Jack Dongarra , Julien Herrmann , Stanimire Tomov

ACM Transactions on Mathematical Software, 2013, 39 (2), pp.1 - 13. ⟨10.1145/2427023.2427025⟩

Article dans une revue hal-00908496v1
Deposit thumbnail

Task-Based Parallel Programming for Scalable Algorithms: application to Matrix Multiplication

Emmanuel Agullo , Alfredo Buttari , Abdou Guermouche , Julien Herrmann , Antoine Jego

[Research Report] RR-9461, Inria Bordeaux - Sud-Ouest. 2022, pp.29

Rapport hal-03588491v2
Deposit thumbnail

Optimal Memory-aware Backpropagation of Deep Join Networks

Olivier Beaumont , Julien Herrmann , Guillaume Pallez , Alena Shilova

[Research Report] RR-9273, Inria. 2019

Rapport (rapport de recherche) hal-02131552v1
Deposit thumbnail

Acyclic partitioning of large directed acyclic graphs

Julien Herrmann , M Yusuf Özkaya , Bora Uçar , Kamer Kaya , Ümit V. Çatalyürek

[Research Report] RR-9163, Inria - Research Centre Grenoble – Rhône-Alpes. 2018

Rapport (rapport de recherche) hal-01744603v1
Deposit thumbnail

A scalable clustering-based task scheduler for homogeneous processors using DAG partitioning

Yusuf M. Özkaya , Anne Benoit , Bora Uçar , Julien Herrmann , Ümit V. Çatalyürek

[Research Report] RR-9185, Inria Grenoble Rhône-Alpes. 2018, pp.1-34

Rapport (rapport de recherche) hal-01817501v3
Deposit thumbnail

Periodicity is Optimal for Offline and Online Multi-Stage Adjoint Computations

Guillaume Aupy , Julien Herrmann

[Research Report] RR-8822, INRIA Grenoble - Rhône-Alpes. 2015

Rapport (rapport de recherche) hal-01244584v1
Deposit thumbnail

Memory-aware list scheduling for hybrid platforms

Julien Herrmann , Loris Marchal , Yves Robert

[Research Report] RR-8461, INRIA. 2014, pp.30

Rapport (rapport de recherche) hal-00944336v1
Deposit thumbnail

Optimal Multistage Algorithm for Adjoint Computation

Guillaume Aupy , Julien Herrmann , Paul Hovland , Yves Robert

[Research Report] RR-8721, LIP - ENS Lyon; Argonne National Laboratory; INRIA; Vanderbilt University. 2014

Rapport (rapport de recherche) hal-01147155v2
Deposit thumbnail

Determining the optimal redistribution

Thomas Hérault , Julien Herrmann , Loris Marchal , Yves Robert

[Research Report] RR-8499, INRIA. 2014

Rapport (rapport de recherche) hal-00960452v2
Deposit thumbnail

Tree traversals with task-memory affinities

Julien Herrmann , Loris Marchal , Yves Robert

[Research Report] RR-8226, INRIA. 2013, pp.31

Rapport (rapport de recherche) hal-00787753v1
Deposit thumbnail

Accelerating linear system solutions using randomization techniques

Marc Baboulin , Jack Dongarra , Julien Herrmann , Stanimire Tomov

[Research Report] RR-7616, INRIA. 2011

Rapport (rapport de recherche) inria-00593306v1
Deposit thumbnail

A scalable clustering-based task scheduler for homogeneous processors using DAG partitioning

Yusuf M. Özkaya , Anne Benoit , Bora Uçar , Julien Herrmann , Ümit V. Çatalyürek

IPDPS 2019 - 33rd IEEE International Parallel & Distributed Processing Symposium, May 2019, Rio de Janeiro, Brazil. pp.155-165, ⟨10.1109/IPDPS.2019.00026⟩

Communication dans un congrès hal-02082794v1
Deposit thumbnail

Acyclic Partitioning of Large Directed Acyclic Graphs

Julien Herrmann , Jonathan Kho , Bora Uçar , Kamer Kaya , Ümit V. Çatalyürek

CCGRID 2017 - 17th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, May 2017, Madrid, Spain. pp.371-380, ⟨10.1109/CCGRID.2017.101⟩

Communication dans un congrès hal-01672010v1
Deposit thumbnail

Computing the expected makespan of task graphs in the presence of silent errors

Henri Casanova , Julien Herrmann , Yves Robert

Ninth International Workshop on Parallel Programming Models and Systems Software for High-End Computing (P2S2), 2016, Aug 2016, Philadelphia, United States

Communication dans un congrès hal-01354711v1
Deposit thumbnail

Bridging the Gap between Performance and Bounds of Cholesky Factorization on Heterogeneous Platforms

Emmanuel Agullo , Olivier Beaumont , Lionel Eyraud-Dubois , Julien Herrmann , Suraj Kumar et al.

Heterogeneity in Computing Workshop 2015, May 2015, Hyderabad, India. ⟨10.1109/IPDPSW.2015.35⟩

Communication dans un congrès hal-01120507v1

Determining the Optimal Redistribution for a Given Data Partition

Thomas Hérault , Julien Herrmann , Loris Marchal , Yves Robert

13th International Symposium on Parallel and Distributed Computing (ISPDC), 2014, Marseille, France. ⟨10.1109/ISPDC.2014.16⟩

Communication dans un congrès hal-01111537v1
Deposit thumbnail

Model and complexity results for tree traversals on hybrid platforms

Julien Herrmann , Loris Marchal , Yves Robert

HeteroPar - International Workshop on Algorithms, Models and Tools for Parallel Computing on Heterogeneous Platforms, Aug 2013, Aachen, Germany

Communication dans un congrès hal-00926502v1
Deposit thumbnail

Designing LU-QR hybrid solvers for performance and stability

Mathieu Faverge , Julien Herrmann , Julien Langou , Bradley Lowery , Yves Robert et al.

IEEE International Parallel & Distributed Processing Symposium (IPDPS 2014), May 2014, Phoenix, United States

Communication dans un congrès hal-00930238v1
Deposit thumbnail

Memory-aware Algorithms and Scheduling Techniques for Matrix Computattions

Julien Herrmann

Distributed, Parallel, and Cluster Computing [cs.DC]. Ecole normale supérieure de lyon - ENS LYON, 2015. English. ⟨NNT : 2015ENSL1043⟩

Thèse tel-01241485v1