Cookies Policy

The website of the University of Cádiz uses its own and third-party cookies to carry out analysis of use and measurement of traffic, as well as to allow the correct functioning in social networks, and in this way to improve your browsing experience.

If you want to configure cookies, press the button Customize Cookies. You can also access the cookie settings at any time from the corresponding link in the footer.

For more information about cookies you can consult the Cookies policy from the website of the University of Cádiz.

Cookies customization

The website of the University of Cádiz uses its own and third-party cookies to carry out analysis of use and measurement of traffic, as well as to allow the correct functioning in social networks, and in this way to improve your browsing experience.

For more information about cookies you can consult the Cookies policy from the website of the University of Cádiz. You can also access the cookie settings at any time from the corresponding link in the footer.

You can configure the website cookies according to their purpose:

  • Statistical analysis

    Third-party cookies (Google Analytics) are used on this site that allow the number of users to be quantified anonymously (personal data will never be obtained to identify the user) and thus be able to analyze the use made by users of our service, in order to improve the browsing experience and offer our content optimally.

  • Social networks

    Third-party cookies are used on this website that allow the proper functioning of some social networks (mainly YouTube and Twitter) without using any personal data of the user.

UniversidaddeCádiz
Instituto de Investigación en Ingeniería y Tecnologías Avanzadas para la Industria Digital -ERICA-

Support System

In addition to the material resources of the ETSI, the UCA has a Cluster of Computers for Research Support (CAI). This supercomputer is composed of 3 HP class-C chassis (c7000), each of which houses 16 gen8 bl460 nodes, resulting in 48 nodes. Each chassis supports the 16 nodes with only 3 power supplies, although now for redundancy it has 6 units. It is also cooled by 10 PARSEC fans. Each node has 2 Intel Xeon E5 2670 processors at 2.6 GHz. Each processor is octo-core, so 48 x 8 x 2 = 768 computing cores are available. The performance is 8 floating point operations per core and cycle (8 flops), with each node working at 2.6 GHz, results in 2.6GHz x 8 flops = 20.8 Gflops, and for the 768 cores it obtains approximately 16 Tflops, that means 16 billion floating operations per second. In addition, each node has 128 Gb RAM, so the Cluster as a whole includes more than 6 TB of main memory. The local storage is made in a SATA disk drive of 500GB of capacity and 10 Krpm. However, global storage, node images, temporary scratch space and user accounts are implemented via an HP P2000 Storage Area Network (SAN). This device has dual RAID controller, and 60 TB gross capacity. The fiber connection to the nodes reaches 8 Gbps.

In relation to the network, three differentiated networks have been designed. The calculation network, inaccessible from the outside, has 10 Gigabit Ethernet architecture with ultra low latency, and allows the exchange of information between the nodes when computing. The main backplane reaches 160 Gbps, connecting each chassis to 40 Gbps. The other console networks and internal management allow the management of the nodes and their cloning and configuration.