The IA datacenter has multiple computational resources. These are mostly associated to researchers and faculty.
In particular, in coordination with the CATA project, there is the Geryon2 computing cluster: a set of 20 servers that can currently offer 792 cores and 7.3 TB of RAM, with approximately 300 TB of storage, of which 56 TB correspond to fast I/0 disks. In addition, there are two GPU Tesla units with a total of 3072 Cores and 200 GB RAM. As a whole, the cluster has a processing capacity of 31 TFlops.
All these equipments are interconnected at 10 Gbps, the internal network works at 1 Gbps and the connection with the rest of the San Joaquín Campus is carried out at 10 Gbps, complemented by the university’s access to the national and international network (through REUNA ) and access via VPN with our own administration.
In September 2023, the IA will house a new cluster consisting of 12 nodes with a total of 768 CPU cores @ 4.1 GHz, 6 TB of RAM, a 10 Gbps Ethernet network and a 100 Gbps IB network, purchased by the Basal Project/CATA 2.
Everything is backed up by a 80 KVa UPS system and a generator with 40 hours of autonomy, which to date has ensured uninterrupted service. In addition, the data center is adequately heated by a dedicated system with a capacity of 400,000 btu.
Each graduate student has access to the network through a computer or notebook made available to them by the program during the time they are in it.
On the IA workstations, in addition to freely available software for the Linux environment (such as compilers, utilities like OpenOffice, Python, etc.), sophisticated packages for data analysis and reduction, including IDL, are regularly installed and maintained. ESO pipelines, Astropy, Anaconda, etc. The University has some educational licenses available, such as Microsoft – Campus Agreement, Office 365, Microsoft – Azure Dev Tools for Teaching, Mathematica, MathWorks/Matlab, Wolfram/Mathematica, Smart Pack (SPSS, Statistical Package for the Social Sciences), etc. High speed wireless network access (including eduroam) is also available. Faculty also have access to professional Zoom licenses. We have a 30TB internal cloud available and we are also developing a local GitLab service and a node bank for Jupyter Lab.
All of these resources are primarily intended for use by the astronomical community. However, the IA also considers collaboration with other research areas that require access to computers to be of great interest. More details can be found in this site.
