When we talk about the states of a microprocessor, what we care about are the contents of the registers, the cache, and the instructions in the pipeline. I'm not certain about the gigahertz level processors of today, but for the processors of 20 years ago, these states are completely stable in terms of changing the placement of one electron with non-relativistic energy levels.

Those processors operated at tens of megahertz. Are 100 Hz neurons so much more sensitive that the placement of a single electron has any effect on the mind states we care about?