The cluster is composed of 83 CPU computing nodes:
- 32 nodes with Dual Intel Xeon Gold 6248 @ 2.5GHz – 20 cores processors, i.e. 40 cores per node and 192 GB of memory.
- 16 nodes with 2 AMD Epyc 7302 @ 3 GHz – 16 cores processors, i.e. 32 cores per node and 256 GB of memory.
- 35 nodes with 2 AMD EPYC 9124 processors, 3 GHz with 16 cores (= 32 cores per machine) and 384 GB RAM.
6 nodes for GPU computing:
- 2 nodes equipped with:
- 4 Nvidia Tesla V100 32 GB GPU cards
- 2 Intel Xeon Gold 5218 @ 2.30GHz x 16 core CPUs
- 384 GB of memory.
- 1 node equipped with:
- 4 Nvidia A100 80 Go GPU cards
- 2 Intel Xeon Gold 5320, 2,2GHz x 26 core CPUs
- 512 GB of memory.
- 3 nodes equipped with:
- 4 NVIDIA H100 NVL cards, PCIe, 350 W-400 W, 94 GB
- 2 Intel Xeon Gold 5420+ 2 GHz processors with 28 cores each
- 512 GB RAM
And also:
- 1 SMP type node with a large memory:
- 4 Intel Xeon Gold 5220 @ 2.2GHz x 18 core CPUs;
- 1.5 TB RAM.
- 1 visualisation node :
- 2 x Intel(R) Xeon(R) Gold 6230R CPU @ 2.10GHz – 26 cores
- 384 GB memory
- Graphic card Quadro RTX 6000
- 80 TB of HomeDir storage (NFS).
- 480TB of scratch storage (BeeGFS).
- InfiniBand HDR 200Gb/s interconnection.