It can be enlightening to look at the number of subdirectories under
/arch in a full-blown Linux source-tree. On my system there are 24. Some, obviously, are close-cousins of other ones,
but nonetheless, we have a system here that can run ... the
same system ... on a multitude of different architectures.
Linux can also control a huge number of CPUs. It can operate in a computing
cluster where CPUs run at differing speeds and are connected in a network that may be fast or slow. This really
is "rocket-science stuff." (As to what it may be used for around the world, I'm afraid that you'll pretty-soon need some gosh-darned-high security clearances to find out a lot of it...)
People who are doing "really big computing" on really-big computers, or really-big
numbers of computers, or both, can find what they're looking for in Linux ... and, frankly, not a whole lot of other places. There's more to the Penguin than just a happy smile.