Yeah the 5 dimensional torus network used at MIRA is just too cool not to bring up here. Modern supercomputers are increasingly become less and less discrete.
If so, that could be interpreted as simply a 4D square grid which wraps around the edges, right? (just as a 3D torus is a 2D grid which wraps)
Also here's an image... which I admit is not terribly useful, but its what the national lab people put out. https://computing.llnl.gov/tutorials/bgq/images/5Dtorus.400p...