High performance computing

Computationally intensive problem solving requires the use of parallel computers and dedicated hardware in a wide range of fields in science, engineering and industry. Many applications need high computational power to keep them moving forward, and for these, high-performance computing is the key tool. High performance computing techniques are essential to extract the full capability offered by different computing platforms, such as multi-core processors, manycore, supercomputers, GPUs, FPGAs, special purpose co-processors (such as tensor computing units) or even future quantum processing units (QPUs). Efficient utilisation and performance improvement of these platforms requires actions at both hardware and software levels. At a higher level, Big Data processing technologies can also be used to solve complex problems in large computer clusters. Of particular interest is also cloud computing, which establishes a new paradigm for service delivery over the network that universalises high-performance computing.

The lines of research in which this programme is framed cover two fundamental aspects, on the one hand, the study and programming of the different types of computing systems mentioned above, and on the other, their use in various fields of application for which solutions are designed based on techniques including those of Artificial Intelligence. The applications in which the greatest effort has been made in recent years are those related to the simulation of electronic devices and solar cells, the processing of hyperspectral and LiDAR data, image processing and bioinformatics, among others.