University of Coimbra
|
Allocation of computing time for scientific or enterprise projects
Hardware 1 x Login and Management node: Fujitsu PRIMERGY RX300 S8 2 Intel Xeon E5-2630v2 CPU (6 core) at 2.60 GHz and a 15MB cache 64 GB of registered and ECC DDR3 LV RAM 2,4 TB on a RAID 5 array (5 x 600 GB SAS at 6 Gbps) 1 RAID controller (SAS 6 Gbps with 1 GB cache) 1 HCA InfiniBand FDR 56 Gbps - 1 port 1 x Lustre MDS 1 Intel Xeon E5-2630v2 CPU (6 core) at 2.60 GHz and a 15MB cache 128 GB of registered and ECC DDR3 LV RAM 1,8 TB on a RAID 10 array (6 x 600 GB SAS at 6 Gbps) 1 RAID controller (SAS 6 Gbps with 1 GB cache) 1 HCA InfiniBand FDR 56 Gbps - 1 port 5 x Lustre OSS 1 Intel Xeon E5-2630v2 CPU (6 core) at 2.60 GHz and a 15MB cache 32 GB of registered and ECC DDR3 LV RAM 300 GB on RAID 10 array (2 x 600 GB SAS at 6 Gbps) 36 TB on a RAID 5 array (10 x 4 TB SATA at 6 Gbps) 1 RAID controller (SAS 6 Gbps with 1 GB cache) 1 HCA InfiniBand FDR 56 Gbps - 1 port 164 x computing nodes in 41 multi-node Fujitsu PRIMERGY CX400 S2 servers 2 Intel Xeon E5-2697v2 (12 core) at 2.70 GHz and a 30 MB cache 96 GB of registered and ECC DDR3 LV RAM 1 HCA InfiniBand FDR 56 Gbps - 1 port Software Operating Software System: CentOS 6.5 (updating to 7.5) Lustre version: 2.5.3 (updating to 2.10.6) Resource Manager: Slurm Scheduler: Slurm Allocation Management: Slurm bank User Environment Manager: Environment Modules Compilers: GNU and Intel