From Grid5000
Jump to: navigation, search

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:

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 2021-05-12 09:17): No current events, None planned (details)

Random pick of publications

Five random publications that benefited from Grid'5000 (at least 2167 overall):

  • David Sierra González, Anshul Paigwar, Özgür Erkent, Jilles Dibangoye, Christian Laugier. Leveraging Dynamic Occupancy Grids for 3D Object Detection in Point Clouds. 16th IEEE International Conference on Control, Automation, Robotics and Vision (ICARCV), Dec 2020, Shenzhen, China. hal-03044979 view on HAL pdf
  • Tom Cornebize, Arnaud Legrand. Simulation-based Optimization and Sensibility Analysis of MPI Applications: Variability Matters. 2021. hal-03141988 view on HAL pdf
  • Luke Bertot, Lucas Nussbaum, David Margery. Implementing SFA Support on an Established HPC-flavored Testbed: Lessons Learned. CNERT 2020 - Computer and Networking Experimental Research using Testbeds, in conjunction with IEEE INFOCOM 2020, Jul 2020, Toronto, Canada. pp.1-6. hal-02962845 view on HAL pdf
  • Neil Ayeb, Eric Rutten, Sebastien Bolle, Thierry Coupaye, Marc Douet. Coordinated autonomic loops for target identification, load and error-aware Device Management for the IoT. FedCSIS 2020 - 15th Federated Conference on Computer Science and Information Systems, Sep 2020, Sofia, Bulgaria. pp.1-10. hal-02934785 view on HAL pdf
  • Romain Serizel, Nicolas Turpault. Sound Event Detection from Partially Annotated Data: Trends and Challenges. IcETRAN conference, Jun 2019, Srebrno Jezero, Serbia. hal-02114652v2 view on HAL pdf

Latest news

Rss.svgUpgrade of the graffiti-13 node in Nancy

We are happy to announce that the graffiti-13 node on the Nancy site in production queue has been upgraded:

its 4 GeForce RTX 2080 Ti 11GB have been replaced by 4 RTX 6000 24GB GDDR6[1].

Reservation example : oarsub -q production -p "host=''" -I


-- Grid'5000 Team 17:00, May 6th 2021 (CEST)

Rss.svgYeti nodes now equipped with 2 NVMe of 1.6 TB and made exotic resources

The yeti nodes of the Grenoble site now have 2 NVMe drives of 1.6TB again (back to their original configuration of 2018) and the troll nodes have new NVMe (1.6TB also, but newer model).

The 4 new NVMe were financed by the Software Heritage project. Many Thanks.

Since the 4 yeti nodes are the only quadri-cpu resources of Grid'5000 and with 2 NVMe, they are now defined as exotic resources, and must be reserved with `-t exotic` in the oarsub arguments.

-- Grid'5000 Team 16:00, April 21th 2021 (CEST)

Rss.svgDocker cache available inside Grid5000

With the growing use of Docker, you might have experienced some rate limiting when pulling images due to the Docker Hub recent policy change.

As a result, we now provide a Docker cache accessible from inside Grid'5000 at the following address

For configuration instructions and more information, see: Using

-- Grid'5000 Team 09:00, April 13th 2021 (CEST)

Rss.svgGrid'5000 environments now use zstd compression

We are moving the default image archive format used for Grid'5000 environments from tar.gz to tar.zst. Switching to zstd allows for significantly faster decompression speed.

This change is transparent for most usages. Kadeploy will use zstd transparently for environments using it (e.g. new environments), the main impact is that deployment will be faster.

If kameleon is used to generate an environment (thus creating an environment description file and an image archive), it will now automatically use zstd by default.

However if you manually modify the description of an environment and/or regenerate an image archive with tgz-g5k (using the -z switch for zstd), you may have to make sure that both use the same compression.

This change is part of a larger effort to improve the performance of kadeploy.

-- Grid'5000 Team 16:00, March 31st 2021 (CEST)

Read more news

Grid'5000 sites

Current funding

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


Logo INRIA.gif




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

Provence Alpes Côte d'Azur
Hauts de France