Grid5000:Home

Grid'5000

Grid'5000 is a large-scale and flexible testbed for experiment-driven research in all areas of computer science, with a focus on parallel and distributed computing including Cloud, HPC and Big Data and AI.

Key features:

provides access to a large amount of resources: 15000 cores, 800 compute-nodes grouped in homogeneous clusters, and featuring various technologies: PMEM, GPU, SSD, NVMe, 10G and 25G Ethernet, Infiniband, Omni-Path
highly reconfigurable and controllable: researchers can experiment with a fully customized software stack thanks to bare-metal deployment features, and can isolate their experiment at the networking layer
advanced monitoring and measurement features for traces collection of networking and power consumption, providing a deep understanding of experiments
designed to support Open Science and reproducible research, with full traceability of infrastructure and software changes on the testbed
a vibrant community of 500+ users supported by a solid technical team

Read more about our teams, our publications, and the usage policy of the testbed. Then get an account, and learn how to use the testbed with our Getting Started tutorial and the rest of our Users portal.

Grid'5000 is merging with FIT to build the SILECS Infrastructure for Large-scale Experimental Computer Science. Read an Introduction to SILECS (April 2018)

Recently published documents and presentations:

Presentation of Grid'5000 (April 2019)
Report from the Grid'5000 Science Advisory Board (2018)

Older documents:

Grid'5000 is supported by a scientific interest group (GIS) hosted by Inria and including CNRS, RENATER and several Universities as well as other organizations. Inria has been supporting Grid'5000 through ADT ALADDIN-G5K (2007-2013), ADT LAPLACE (2014-2016), and IPL HEMERA (2010-2014).

Current status (at 2025-12-28 09:58): 6 current events, 2 planned (details)

Random pick of publications

Five random publications that benefited from Grid'5000 (at least 2928 overall):

Reda Khoufache, Anisse Belhadj, Hanene Azzag, Mustapha Lebbah. Distributed MCMC inference for Bayesian Non-Parametric Latent Block Model. Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD), May 2024, Taipei, Taiwan. hal-04457575 view on HAL pdf
Jolan Philippe, Antoine Omond, Hélène Coullon, Charles Prud'Homme, Issam Raïs. Fast Choreography of Cross-DevOps Reconfiguration with Ballet: A Multi-Site OpenStack Case Study. SANER 2024: IEEE International Conference on Software Analysis, Evolution and Reengineering, Mar 2024, Rovaniemi, Finland. pp.1-11, 10.1109/SANER60148.2024.00007. hal-04457484 view on HAL pdf
Pierre Epron, Gaël Guibon, Miguel Couceiro. ORPAILLEUR SyNaLP at CLEF 2024 Task 2: Good Old Cross Validation for Large Language Models Yields the Best Humorous Detection. Working Notes of the Conference and Labs of the Evaluation Forum (CLEF 2024), Sep 2024, Grenoble, France. pp.1841-1856. hal-04696012 view on HAL pdf
Guillaume Helbecque, Ezhilmathi Krishnasamy, Nouredine Melab, Pascal Bouvry. GPU-Accelerated Tree-Search in Chapel versus CUDA and HIP. 14th IEEE Workshop Parallel / Distributed Combinatorics and Optimization (PDCO 2024), May 2024, San Francisco, United States. 10.1109/IPDPSW63119.2024.00156. hal-04551856 view on HAL pdf
Eva Giboulot, Teddy Furon. WaterMax: breaking the LLM watermark detectability-robustness-quality trade-off. NeurIPS 2024 - 38th Conference on Neural Information Processing Systems, Dec 2024, Vancouver, Canada. pp.1-34. hal-04766606 view on HAL pdf

Latest news

End of support for centOS7/8 and centOSStream8 environments

Support for the centOS7/8 and centOSStream8 kadeploy environments is stopped due to the end of upstream support and compatibility issues with recent hardware.

The last version of the centOS7 environments (version 2024071117), centOS8 environments (version 2024071119), centOSStream8 environments (version 2024070316) will remain available on /grid5000. Older versions can still be accessed in the archive directory (see /grid5000/README.unmaintained-envs for more information).

-- Grid'5000 Team 08:44, 4 December 2025 (CEST)

Ecotaxe cluster is now in default queue at Nantes

We are pleased to announce that the ecotaxe cluster of Nantes is now available in the default queue.

As a reminder, ecotaxe is a cluster composed of 2 HPE ProLiant DL385 Gen10 Plus v2 servers[1].

Each node features:

2 AMD EPYC 7453 (Zen 3), 28 cores/CPU

3 Nvidia A100 80GB GPU

256 GB memory

1x 1.92 To SSD + 2x 7.68 To SSD

100 Gb/s Intel Ethernet adapter [2].

To submit a job on this cluster, the following command may be used:

oarsub -t exotic -p ecotaxe

This cluster is co-funded by Région Pays de la Loire, FEDER and REACT EU via the CPER SAMURAI [3].

[1] https://www.grid5000.fr/w/Nantes:Hardware#ecotaxe

[2] The observed throughput depends on multiple parameters such as the workload, the number of streams, ... [3] https://www.imt-atlantique.fr/fr/recherche-innovation/collaborer/projet/samurai

-- Grid'5000 Team 14:10, 02 December 2025 (CET)

Some changes on the hardware configuration of Grenoble nodes

We recently did some hardware changes on clusters yeti, troll and dahu.

The changes are as follows:

yeti :

Following a malfunction of the two NVMe disks on yeti-3, an NVMe disk from yeti-1 has been transferred to yeti-3 to ensure that we have at least one functional NVMe disk per yeti node. New NVMe configuration of the nodes:

yeti-[1,3]: 1× NVMe
yeti-[2,4]: 2× NVMe

troll :

Due to experimentation needs, the steering committee agreed to change the hardware configuration of the troll cluster, replacing the Omnipath HPC network interconnect (interconnecting troll to yeti and dahu) by the Infiniband HPC network interconnect already available for the drac cluster.

dahu :

A few nodes of the dahu cluster recently encountered a recurrent problem with their OPA interfaces. Instead of fully retiring those nodes, we chose to disable their OPA interfaces.

This change means that if you want to reserve a dahu node with OPA, you must specify it in your oarsub request. For example:

oarsub -I -p "dahu and opa_count > 0"

The nodes that have been modified are dahu-18, dahu-26 and dahu-30. More nodes may be added to this list in the future.

-- Grid'5000 Team 14:50, 24 November 2025 (CEST)

Cluster "clervaux" is now in the default queue in Luxembourg

We are pleased to announce that the clervaux[1] cluster of Luxembourg is now available in the default queue.

Clervaux is a cluster composed of 48 CPU nodes.

Each node features:

2x CPU Intel Xeon E5-2680 v4 (14 cores/CPU, 2 threads/cores)

128 GiB RAM

1x 120GB SSD SATA disk

This cluster was funded by the University of Luxembourg.

[1] https://www.grid5000.fr/w/Luxembourg:Hardware#clervaux

-- Grid'5000 Team 10:50, 21 October 2025 (CEST)

Grid'5000 sites

Current funding

As from June 2008, Inria is the main contributor to Grid'5000 funding.

INRIA

CNRS

Universities

IMT Atlantique
Université Grenoble Alpes, Grenoble INP
Université Rennes 1, Rennes
Institut National Polytechnique de Toulouse / INSA / FERIA / Université Paul Sabatier, Toulouse
Université Bordeaux 1, Bordeaux
Université Lille 1, Lille
École Normale Supérieure, Lyon

Regional councils

Aquitaine
Auvergne-Rhône-Alpes
Bretagne
Champagne-Ardenne
Provence Alpes Côte d'Azur
Hauts de France
Lorraine