Our computing clusters
CPU-ALF - Clúster CPU de computación de alto rendimiento del grupo de investigación en Aplicaciones del Láser y Fotónica de la USAL
The ALF-USAL group has a high-performance computing cluster that has been essential for performing local numerical simulations of laser-matter interaction. The cluster was installed in 2017 as a result of a Leonardo Grant project from the BBVA Foundation and has grown thanks to the purchase of equipment through various research projects.
Currently, it allows for high-performance parallelized computing on both CPUs (Central Processing Units) and GPUs (Graphical Processing Units). Although the CPU computing capacity is significantly greater, the GPUs have enabled the first laser-matter interaction simulation calculations using artificial intelligence.
The equipment details are as follows:
- 12 CPU computing nodes. Each has 2 Intel Xeon Gold processors, 192 GB of RAM, 56G InfiniBand connectivity, and 10G Ethernet connectivity. All nodes share 24 GB of storage protected with RAID-5 redundancy.
- 4 GPU computing cards: three NVIDIA A30 cards, suitable for any scientific computing (including double precision), and one NVIDIA RTX 3080 card, suitable for single-precision computing (such as deep learning).
- Synology NAS for network data storage with eight 16 TB hard drives, protected with RAID-5 redundancy.
GPU-SINAMAG - Clúster GPU de computación de alto rendimiento del grupo de investigación de Nanosestructuras Magnéticas de la USAL
The SINAMAG group has a GPU (Graphics Processing Units) cluster for numerical computing. The cluster began construction in 2012 and has since grown steadily, increasing by approximately 20 TFlops per year. This growth and the upgrade of the equipment has been made possible
thanks to the funding obtained by the group for research projects, both at the national and regional levels.
Current computing capacity exceeds 400 Tflops, while storage capacity is around 200 TB.