Grid5000:Home
|
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:
Grid'5000 is merging with FIT to build the SILECS Infrastructure for Large-scale Experimental Computer Science. Read an Introduction to SILECS (April 2018)
Older documents:
|
Random pick of publications
Five random publications that benefited from Grid'5000 (at least 2758 overall):
- Shashikant Ilager, Daniel Balouek, Sidi Mohammed Kaddour, Ivona Brandic. Proteus: Towards Intent-driven Automated Resource Management Framework for Edge Sensor Nodes. FlexScience'24: Proceedings of the 14th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, Jun 2024, Pisa, Italy. pp.1 - 8, 10.1145/3659995.3660037. hal-04775138 view on HAL pdf
- Miguel Couceiro, Esteban Marquer, Pierre-Alexandre Murena, Pierre Monnin. Workshop Analogies: from Theory to Applications (ATA@ICCBR 2022). Workshop Analogies: from Theory to Applications (ATA@ICCBR 2022), CEUR Workshop Proceedings, 3389, CEUR-WS.org, pp.3-103, 2023, Workshop Proceedings of the 30th International Conference on Case-Based Reasoning co-located with the 30th International Conference on Case-Based Reasoning (ICCBR 2022). hal-04392022 view on HAL pdf
- Berné Nortier, Mostafa Sadeghi, Romain Serizel. Unsupervised speech enhancement with diffusion-based generative models. International Conference on Acoustics Speech and Signal Processing (ICASSP), IEEE, Apr 2024, Seoul (Korea), South Korea. 10.48550/arXiv.2309.10450. hal-04210707v2 view on HAL pdf
- Melvin Chelli, Cédric Prigent, René Schubotz, Alexandru Costan, Gabriel Antoniu, et al.. FedGuard: Selective Parameter Aggregation for Poisoning Attack Mitigation in Federated Learning. Cluster 2023 - IEEE International Conference on Cluster Computing, Oct 2023, Santa Fe, New Mexico, United States. pp.1-10, 10.1109/CLUSTER52292.2023.00014. hal-04208787 view on HAL pdf
- Hammouda Elbez, Mazdak Fatahi. Progressive Layer-based Compression for Convolutional Spiking Neural Network. 2023. hal-03826823v2 view on HAL pdf
Latest news
Change of default queue based on platform
Until now, Abaca (production) users had to specify `-q production` when reserving Abaca resources with OAR.
This is no longer necessary as your default queue is now automatically selected based on the platform your default group is associated to, as shown at https://api.grid5000.fr/explorer/selector/ and in the message displayed when connecting to a frontend.
For SLICES-FR users, there is no change since the correct queue was already selected by default.
Additionally, the "production" queue has been renamed to "abaca", although "production" will continue to work for the foreseeable future.
Please note one case where this change may affect your workflow:
When an Abaca user reserves a resource from SLICES-FR (a non-production resource), they must explicitly specify they want to use the SLICES-FR queue, which is called "default", by adding `-q default` the OAR command.
-- Abaca Grid'5000 Team 10:10, 31 March 2025 (CEST)
Cluster "musa" with Nvidia H100 GPUs is available in production queue
We are pleased to announce that a new cluster named "musa" is available in the production queue¹ of Abaca.
This cluster has been funded by Inria DSI as a shared computing resource.
It is accessible to all Abaca users. Users affiliated with Inria have access with the same level of priority, regardless of the research center to which they are attached.
This cluster is composed of six HPE Proliant DL385 Gen11 nodes² with 2 AMD EPYC 9254 24-Core Processor, 512 GiB of RAM, 2 x Nvidia H100 NVL (94 GiB) with NVLink, one 6 TB SSD NVME and 25 Gbps Ethernet Connexion
Please note that in order to share it efficiently, walltime is limited:
The cluster "musa" is located at Sophia, hosted in the datacenter of Inria Centre at Université Côte d’Azur.
¹: https://api.grid5000.fr/explorer/hardware/sophia/#musa
²: the nodes are named musa-1, musa-2,.., musa-6
-- Grid'5000 Team 13:30, 19 March 2025 (CEST)
Cluster "Hydra" is now in the testing queue in Lyon
We are pleased to announce that the hydra[1] cluster of Lyon is now available in the testing queue.
Hydra is a cluster composed of 4 NVIDIA Grace-Hopper servers[2].
Each node features:
Due to its bleeding edge hardware, usual Grid'5000 environments are not supported by default for this cluster.
(Hydra requires system environments featuring a Linux kernel >= 6.6). The default system on the hydra nodes is based on Debian11, but **does not provide functional GPU**. However, users may deploy the ubuntugh2404-arm64-big environment, which is similar to official Nvidia image provided for this machine and provides GPU support.
To submit a job on this cluster, the following command may be used:
oarsub -q testing -t exotic -p hydra
This cluster is funded by INRIA and by Laboratoire de l'Informatique du Parallélisme with ENS Lyon support.
[1] Hydra is the largest of the modern constellations according to Wikipedia: https://en.wikipedia.org/wiki/Hydra_(constellation)
[2] https://developer.nvidia.com/blog/nvidia-grace-hopper-superchip-architecture-in-depth/
-- Grid'5000 Team 16:10, 11 March 2025 (CEST)
New Access Rules for Production Queue
We are introducing new access rules for clusters in the production queue. Clusters in this queue are now accessed based on priority levels that reflect their funding sources. Jobs submitted at higher priority levels are scheduled before those at lower levels and may also have longer maximum durations.
For more detailed information, please visit: https://www.grid5000.fr/w/Production#Using_production_resources . However, an important thing to note is that job submission commands that worked previously will still work after that change (but you might get higher or lower priority depending on your resources selection).
You can check your priority level for each cluster using https://api.grid5000.fr/explorer . Currently, this tool only displays basic information; however, we plan to add more features soon.
Please be aware that these changes apply only to the production clusters, which are currently available only in Nancy and Rennes. There are no changes to the "default" queue.
If you encounter any issues or have feedback regarding this new feature, or if you believe your priority level on specific resources is not adequate, please contact us at <support-staff@lists.grid5000.fr>.
-- Grid'5000 Team 9:25, June 4th 2024 (CEST)
Grid'5000 sites
Current funding
As from June 2008, Inria is the main contributor to Grid'5000 funding.
INRIA
|
CNRS
|
UniversitiesIMT Atlantique |
Regional councilsAquitaine |
