Is that really his name? Word was he was an acerbic, bull-headed drunk. Then the usual idea of computer architecture is just instructions on the left data on the right.
What I think von Neumann has as interesting, besides stuff like "Goedel/Bernays/vonNeumann" accounts of set and class theory or "ZF with classes" after ZFC with GBN, is: that there are accounts of the equi-decomposability in the planar, where Vitali makes the account for the line segment, and Vitali and Hausdorff make the account for the ball, that these days is usually called Banach-Tarski though that Vitali-Hausdorff already did it, then there's that von Neumann at least got into some interesting cases about "Zeno's graduation course: the thought experiment they don't teach you in school", in the planar, or as about the bee's travel between two oncoming trains, that von Neumann has some at least interesting "non-standard accounts", sort of like Peano has a theory of infinitesimals besides a theory of integers, then though mostly he's referenced for reductionism.
1-D linear: one equi-decomposability 2-D planar: multiple equi-decomposabilities 3-D volumetric: one equi-decomposabilitySo, it's interesting for measure theory if he ever invented anything and wasn't just describing what was already there.