It's hard to rule anything out, but I think using non-atomic matter would be very difficult for a few reasons. (1) Subatomic particles are not necessarily smal

It's hard to rule anything out, but I think using non-atomic matter would be very difficult for a few reasons. (1) Subatomic particles are not necessarily smaller than atoms. Even though in the particle sense, an electron is infinitely small, in the wave sense, an electron can extend over a volume much larger than a single atom. So it's not necessarily the case that a subatomic particle like an electron is smaller than an atom. (Also, the positive charge of the nucleus actually helps shrink the electron's range. So adding positive particles can end up making the system smaller.) (2) If you want a transistor every 1.5 nm, then you probably need at least one particle every 1.5 nm. And the only way you can pack matter that densely is if it's charge neutral (a clump of singly charged particles would immediately disperse from electric repulsion). So you either need a combination of positive and negative particles (atoms), or neutral particles like photons/neutrons/neutrinos. The problem with neutral particles is that there's no long-range forces to keep them in whatever structure you design. Gravity is too weak at that scale, and electromagnetism has no effect because they are neutral. Perhaps photons could be used for computation, but without using atoms for mirrors/lenses/waveguides/etc it's hard for me to imagine how. TL;DR: Atoms are an efficient way to pack particles and preserve structure. It's hard to imagine doing it with a plasma or gas or something else.