The performance bottleneck of physics based animation is often the collision detection. It is well known by practitioners
that the collision detection may consume more than half of the simulation time. In this work, we will
introduce a novel approach for collision detection using bounding volume hierarchies. Our approach makes it
possible to perform non-convex object versus non-convex object collision on the GPU, using tandem traversals
of bounding volume hierarchies. Prior work only supports single traversals on GPUs. We introduce a blocked
hierarchy data structure, using imaginary nodes and a simultaneous descend in the tandem traversal. The data
structure design and traversal are highly specialized for exploiting the parallel threads in the NVIDIA GPUs.
As proof-of-concept we demonstrate a GPU implementation for a multibody dynamics simulation, showing an
approximate speedup factor of up to 8 compared to a CPU implementation.