The Cassiopée computing center is the high-performance computing platform of the École Nationale SupérieureArts et Métiers.
Open to all laboratories and students at the institution, itreduces calculation times and enables calculations that cannot be performed on a conventional workstation.
Cassiopée's objective is to provide a high-performance computing service to researchers in the iC-ARTS network laboratories and to students at the institution. Pooling and sharing resources makes it possible to have powerful computing resources at your disposal.
To access computing resources, simply send a request to cassiopee@ensam.eu.
The project team led byArts et Métiers DGARI
- Boris PIOTROWSKI - Project Manager - Scientific Computing Research Engineer
- Jérôme MONELLE -IT Administrator
- Camille BURTON- Scientific Manager - Scientific Computing Research Engineer
- Paolo ERRANTE - Research Engineer
- 250 users spread across 11 research laboratories
Use cases
- Reduction in the time required for a single calculation by executing it in parallel on multiple processing cores (up to several hundred cores simultaneously).
- Increase in the size of simulated models by adding the RAM of several computing nodes (more than 7 TB of RAM),
- Increase in the number of calculations performed simultaneously
- Post-processing of large models
User profiles
- Basic research,
- Educational projects,
- Partnership-based research,
- Industrial partnership.
Laboratories with access to the platform
Updated: March 6, 2025
High-performance computing support services
- Initial training: lasting 2 to 4 hours, this is essential for accessing computing equipment.
- Training update: for occasional users (~30 minutes),
- Software installation (requires a link to a license server for non-free software),
- Assistance with software compilation and optimization tailored to Cassiopée's hardware resources and compilers.
- Code parallelization assistance: optimization of hardware resource usage or assistance with code optimization (parallelization performance diagnostics).
Some additional information for further reading...
The equipment
- 2,500 computing cores,
- 1 TB storage (high performance)
- Infiniband low-latency network
Infrastructure
- HPC gray room: redundant air conditioning
- Capacity of 6 cabinets (200 racks)
- Room supervision
- Information system security
Scientific resources
- Bringing together data scientists around a common platform
- Discussions on digital methods, CASSIOPEE Campus
- Software sharing
Some illustrations...
Parallel Computation of Hierarchical Poromechanical Models Using the Open-Source Software FEniCSx
Cumulative plastic deformation in a metallic polycrystal subjected to simple shear for different grain sizes
Reference: Mareau, Continuum Mechanics and Thermodynamics, 2022.
Contacts
Email: cassiopee@ensam.eu /Address: 4, rue Augustin Fresnel 57078 Metz - Cedex 3 - France
