... were big boxes. I, personally, consider the PDP-8 the start of it all. The PDP-8 (1' * 2' * 2') for all practical purposes had every characteristic of a single chip microcontroller save the size. I remember comparing the 8048 and the PDP-8 for an application (the PDP-8 won).
My first "single chip" project used a 8085 for processing and a homemade lightning fast (for those days) "nanocomputer" (the 4 instructions were: skip if - jmp - set - clear) for the parts that required speed.

Erik