IFIC opens its computer infrastructure for Artificial Intelligence
- Science Park
- February 6th, 2020
The Institute of Corpuscular Physics opens its computer infrastructure for the first time for Artificial Intelligence, Artemisa, to research projects of the Universitat de València and CSIC, in the Valencian Community, that demand Machine Learning or Big Data techniques for their development. This special infrastructure is funded by the European Union and the Valencian government.
The Institute of Corpuscular Physics (IFIC), a joint centre of the Spanish Research Council (CSIC) and the Universitat de València, opens its computer infrastructure for the first time for Artificial Intelligence, Artemisa, to research projects of these scientific institutions in the Valencian Community. During February, researchers can send project proposals that demand techniques such as Machine Learning or Big Data for their development. This special infrastructure is funded by the European Union and the Valencian government in the Operative Program FEDER of the Valencian Community 2014-2020 for the acquisition of infrastructure and equipment of R&D.
Artemisa (ARTificial Environment for ML and Innovation in Scientific Advanced Computing) uses the experience and abilities of its computing centre, which houses one of the nodes devoted to the computer network (Grid) of the data obtained by the ATLAS experiment of the Large Hadron Collider (LHC) from CERN, the largest particle accelerator in the world. By the end of 2018, the Universitat de València started the process of acquisition of the necessary machines for Artemisa, with total funds amounting to a million euros from European funds FEDER and the Valencian Department of Education, Research, Culture and Sports.
This high-performance computing system is used to develop Machine Learning and Big Data Analytics. Machine Learning refers to the most used set of Artificial Intelligence techniques nowadays, since it allows computers to carry out tasks without being previously programmed. Big Data Analytics refers to the set of techniques to handle large amounts of information. Achieving their full potential requires a type of graphics processor called GPU, which extend the abilities of traditional computers’ CPUs.
Artificial Intelligence techniques are revolutionizing the society, as well as the world of research. In particle physics Artificial Intelligence is used to accelerate and improve the research of ‘new physics’ among the vast quantity of collisions produced in the LHC and other great experiments in astroparticle physics and cosmology. These techniques can also be used in Medical Physics, where Artificial Intelligence helps in the diagnosis through algorithms that contribute to identify injuries or tumours. The applications of Artificial Intelligence to solve problems in other areas of science are numerous.
In this context, the IFIC’s infrastructure devoted to Artificial Intelligence turns into an important tool to foster the R&D in the Valencian Community. Artemisa has currently 240 terabytes (TB) for data storage, 22 high-performance computing servers and a total of 8.25 TB RAM, in addition to a powerful computing server equipped with 4 GPUs with high-speed interconnection After a trial period reserved to IFIC’s research projects, in this second phase Artemisa is open to projects from the Universitat de València and CSIC centres in the Valencian Community that need this computing powers. Proposals can be sent upon 29 February.
Among the IFIC’s projects that use this infrastructure are the improvement of medical diagnostic techniques; Machine Learning to identify particles in the experiments of the LHC; as well as in other nuclear physics experiments; and Artificial Intelligence for neutrino physics experiments. Some of these projects are supported by the IFIC’s Scientific Unit for Business Innovation, a tool funded by the Valencian Agency of Innovation (AVI) of the government.
More information:
Categories: Física , Astronomía y Astrofísica , Ciencias Tecnológicas , Ciencias de la Tierra y el Espacio
