Cluster description

Hardware

Compute nodes

Node Entity Type CPUs Memory GPUs
cpu-node[32-39] RPBS Supermicro X8DTT-H 12 (2x Intel Xeon CPU X5650 @ 2.67GHz) 48GB
cpu-node[69-72] Epigenomique Dell PowerEdge FC630 32 (2x Intel Xeon CPU E5-2630 v3 @ 2.40GHz 64GB
cpu-node79 RPBS Dell PowerEdge FC630 56 (2x Intel Xeon CPU E5-2660 v4 @ 2.00GHz) 56GB
cpu-node80 RPBS Dell PowerEdge FC640 96 (2x Intel Xeon Gold 5220R @ 2.20GHz) 192GB
cpu-node[81-84] RPBS Dell PowerEdge FC640 64 (2x Intel Xeon Gold 5218 @ 2.30GHz) 256GB
cpu-node110 RPBS Dell PowerEdge R920 120 (4x Intel Xeon E7-4870 v2 @ 2.30GHz) 512GB
cpu-node[130-145] iPOP-UP Dell PowerEdge C6525 128 (2x AMD EPYC 7452 @ 2.35GHz) 256GB
gpu-node2 CMPLI Asus Z10PE-D8 WS 16 (2x Intel Xeon E5-2623 v3 @ 3.00GHz) 16GB 2x GeForce GTX1070
gpu-node4 CMPLI Asus Z10PE-D8 WS 40 (2x Intel Xeon E5-2630 v4 @ 2.20GHz) 32GB 2x Geforce GTX1080Ti
gpu-node[5-6] RPBS Asus Z10PE-D8 WS 40 (2x Intel Xeon E5-2630 v4 @ 2.20GHz) 32GB 2x GeForce RTX2080Ti
gpu-node[7-8] CMPLI Asus Z10PE-D8 WS 40 (2x Intel Xeon E5-2630 v4 @ 2.20GHz) 32GB 2x GeForce RTX2080Ti
gpu-node9 CMPLI Dell PowerEdge R7525 64 (2x AMD EPYC 7282 @ 2.80GHz) 128GB 3x Quadro RTX6000
gpu-node10 Master BI Dell PowerEdge R7525 32 (2x AMD EPYC 7262 @ 3.20GHz) 64GB 3x Quadro RTX6000
gpu-node11 RPBS Dell PowerEdge T640 32 (2x Intel Xeon Silver 4208 @ 2.10GHz) 96GB 1x GeForce RTX3090
gpu-node13 RPBS Dell PowerEdge T640 32 (2x Intel Xeon Silver 4208 @ 2.10GHz) 96GB 2x GeForce RTX3080Ti
gpu-node14 CMPLI Dell Precision 7920 64 (2x Intel Xeon Gold 6226R @ 2.90GHz) 192GB 3x Quadro RTXA6000
gpu-node15 iPOP-UP ProLiant DL385 Gen10 Plus v2 32 (2x AMD EPYC 7313 @ 3.0GHz) 512GB 2x Tesla A100 80GB
gpu-node16 CMPLI ProLiant DL385 Gen10 Plus v2 32 (2x AMD EPYC 7313 @ 3.0GHz) 512GB 2x Tesla A100 80GB
gpu-node17 iPOP-UP ProLiant DL385 Gen10 Plus v2 32 (2x AMD EPYC 7313 @ 3.0GHz) 512GB 2x Tesla A100 80GB
gpu-node18 iPOP-UP ProLiant DL385 Gen10 Plus v2 32 (2x AMD EPYC 7313 @ 3.0GHz) 512GB 2x Tesla A100 80GB

Infrastructure

Node Entity Type Role
master1 iPOP-UP Dell PowerEdge R440 Slurm controler, Slurm database, gateway
virtserv5 RPBS Dell PowerEdge R440 Virtualization
virtserv6 iPOP-UP HP ProLiant DL380 Gen10 Virtualization
directory CMPLI HP ProLiant DL160 Gen10 LDAP server, DHCP, DNS

Storage

Node Entity Type Storage Role
hot-storage-mds1 iPOP-UP Dell PowerEdge R740xd 8x 1.92TB SSDs Metadata / management server (Lustre)
hot-storage-oss[1-2] iPOP-UP DELL PowerEdge R740xd 20x 3.84TB SSDs Object Storage server (Lustre)
cold-storage1 CMPLI DELL PowerEdge R740xd 12x 12TB HDD NFS server
cold-storage1 (extension) iPOP-UP Dell PowerVault MD1400 12x 12TB HDD NFS server (extension)

Network

Node Entity Type Role Description
cmpli-rtr1 iPOP-UP Juniper SRX1500 Routing, filtering 16x 1GbE + 4x 10GbE ports
cmpli-rtr2 CMPLI Juniper SRX1500 Routing, filtering 16x 1GbE + 4x 10GbE ports
sw-data-mgmt-[1-2] RPBS Juniper EX3300 Management, data 24x 1000BASE-T 1GbE ports
sw-dell-10g-[1,3] RPBS Dell N4032F Data 24x SFP+ 10GbE ports
sw-dell-10g-2 RPBS Dell S4128F-ON Data 28x SFP+ 10GbE ports
sw-dell-100g RPBS Dell S5232F-ON Data 32x QSFP28 100GbE

Software layer

The cluster is managed by Slurm (version 20.11.8).

Scientific software and tools are available through Environment Modules and are mainly based on Conda packages or Singularity images.

Operating System: CentOS 7

Around the cluster management: Proxmox VE

Deployment and configuration are powered by Saltstack, Terraform and Gitlab.