Serial, OptimizedSerial, OpenMP, CUDA and CUDA for multiple GPUs support! The current solver supports all the mentioned backends. The optimized Backends (OptimizedSerial, OpenMP, CUDA and CUDA for multiple GPUs) only support HLL+MUSCL+Constant Bottom Reconstruction. Currently the CMakeLists.txt and the corresponding bash scripts are for leo3e and a training cluster with NVIDIA GPUs of the University of Innsbruck. For different clusters, those files must probably be adapted.