The university education in parallel and high-performance computing often suffers from a significant gap between the effects and potential performance taught in the lectures on the one hand and those practically experienced in exercises or lab courses on the other hand. With a small number of processors, the results obtained are often hardly convincing; however, machines crunching numbers at least a bit are rarely accessible to students doing their first steps in parallel programming. In this contribution, we present our experiences of how a state-of-the- art mid-size Linux cluster (64 dual-board P4 nodes with InfiniBand 4x networking, providing an HPL benchmark performance of almost 0.6 TFlops), bought and operated on a department level primarily for edu- cation and algorithm development purposes, can be used for teaching a large variety of HPC aspects such as basics of parallel algorithms, classi- cal tuning, or hardware-aware programming. Special focus is put on the effects of such an approach on the intensity and sustainability of learning.