Researchers in RUES and Team Legato recently integrated their own dedicated machine, gaia-80, into University of Luxembourg’s Gaia Cluster. The machine has eight 15-core Ivy Bridge-generation Xeon Processors connected to 3 terabytes of RAM and 12 terabytes of fast local storage. This new platform enhances UL’s existing HPC infrastructure by providing a robust platform for processing and visualizing large datasets, running shared memory parallel codes and providing a repeatable environment for performance optimization.
Early tests show that this machine can scale up to problems with around half a billion unknowns for finite element diffusion type problems with total runtimes of around 10 minutes.
This purchase of this machine was jointly funded by:
- FSI-HARVEST, Numerical modelling of smart energy harvesting devices driven by flow-induced vibrations, Marie Curie Career Integration Grants under call FP7-PEOPLE-2012-CIG.
- University of Luxembourg Computational Mechanics Lab Budget.
- Geophysics Group Lab Budget.