At present, in high-performance computing there is a tendency towards the use of architectures of many additional cores as computational support. This has led in recent years to new computing paradigms, very different from the classic, such as computing on GPU and FPGAs.The whole of the three types of computation is what is called heterogeneous computing. This type of equipment opens the doors to the reprogrammable calculation in biomedical engineering, structural calculus, multiphysics simulation, Deep Learning and computer architecture design._x000D_ _x000D_ it would be formed by:_x000D_ 1.-HPC computational system with CPUs supported by GPUs Tesla V100 for heterogeneous CPU/GPU calculation and simulations with requirements in the order of several TB of memory and more than 64 cores in shared memory._x000D_ 2.-Computational system with GPU RTX for Deep Learning_x000D_ 3.-HPC computational system with CPUs and FPGAs Xilinx Alveo 280_x000D_ 4.-Storage System in NVMe of data obtained in the different simulations/calculations_x000D_ 5.-High performance distributed network subsystem to interconnect the different computational systems._x000D_ _x000D_ This equipment comes to give solution to the request of multiple groups of the Institute of Research in Engineering of Aragon, and to several external institutions such as the group of computational hydraulics or the group of computation and physics of complex systems.