CiTIUS computing infrastructure is used to provide computational support for the daily operation of the center, as well as for research in high performance computing and development of projects in collaboration with the industry. The CiTIUS building features a modern CPD ready for rack and blade servers.
These computational resources allow that CiTIUS has the capacity of developing projects and services in the following lines:
- Parallelization of applications
- Optimizing applications on multicore architectures
- Optimizing applications using GPGPU computing
- Design and development of cloud architectures
High Performance Computing Services
HPC Cluster
CiTIUS has a heterogeneous computing cluster at disposal of our researchers, in which they can execute their HPC applications. This cluster has 18 computing nodes:
- 8 HP Proliant BL685c G7 nodes, each with 4 AMD Opteron 6262 HE (16 cores) processors and 128 GB of RAM.
- 5 Dell PowerEdge M910 nodes, each with 4 Intel Xeon L7555 (8 cores, 16 threads) processors and 64 GB of RAM.
- 5 Dell PowerEdge M620 nodes, each with 2 Intel Xeon E5-2650L (8 cores, 16 threads) processors and 64 GB of RAM.
- All computational nodes are connected throughout 10 GbE networks.
This cluster has a storage space above 6 TB, accessible to all cluster nodes through a shared file system. Work management is performed through the queuing system PBS/Torque.
Currently, the software installed on this system includes the GNU/GCC compiler collection, JAVA, Python, the R statistical package, MATLAB, different MPI libraries and several commercial simulators of electronic devices.
GPU Servers
For those researchers that use GPUs, CiTIUS has several servers equipped this type of systems, dedicated to optimizing the execution of their programs as well as the development of applications under the GPGPU programming paradigm.
- Supermicro X8DTG-D server (1 Intel® Xeon® CPU E5520 @ 2.27GHz, 12 GB RAM, Ubuntu 12.04)
- 2 Nvidia GF100 [Tesla S2050]
- 1 Matrox MGA G200W WPCM450
- Dell Precision R5400 server (1 Intel® Xeon® CPU E5540 @ 2.83GHz, 8 GB RAM, Ubuntu 14.04)
- 2 Nvidia G98 [Quadro NVS 295]
- 1 Nvidia GF110 [Geforce GTX 580]
- Dell PowerEdge R720 server(1 Intel® Xeon® CPU E5-2069 @ 2.5GHz, 16 GB RAM, Ubuntu 14.04) with an external PCIexpress enclosure Netstor NA265A
- 1 Nvidia GeForce GTX Titan [Kepler GK110] 6GB
- 1 Nvidia GeForce GTX Titan X [Pascal] 12GB
Parallel computing
CITIUS owns a server equipped with a Xeon Phi 7120P card with 61 cores.IaaS Cloud Services
The Center provides cloud resources for researchers. In particular, they can create their own virtual machines, Linux or Windows, over a private Infrastructure as a Service (IaaS). CiTIUS researchers have complete control over these virtual machines, and therefore this service is suitable for the installation of support services, such as web servers or application servers, as well as for creating development or testing environments.
Additionally, this cloud infrastructure promotes teamworking, since it allows the creation of working groups on projects with shared resources.
The cloud is made up of 5 Dell PowerEdge M620 nodes, each equipped with 2 Intel Xeon E5-2650L processors (8 cores, 16 threads) y 64 GB of RAM.
Management Services
The Infrastructure unity supports researchers in order to provide a suitable environment for the development of their research:
- Custom Ubuntu desktop distribution.
- Remote profiles with automatic backups.
- Shared file systems.
- Project management software and hosting with version control.
