The ever increasing compute capacity of high performance computing (HPC) systems enables scientists to simulate and explore physical phenomena with an enormous spatial and temporal accuracy. On the other hand, this accuracy leads to datasets of many terabytes, petabytes, and even exabytes envisioning the up- coming exascale area projected for 2018. To understand complex physical coherences behind such a simulation, an efficient analysis and visualization is essential but also difficult, since the challenges concern all stages of the visual- ization pipeline. With this presentation we set the focus on distributed and hybrid rendering.