How is data physically held in computer memory?

2 bits equal a shave and a hair cut.

The politically correct word for a collection of eight bytes treated together is octet, so as to avoid any size-ist missunderstanding.

The prefered spelling of a four-bit object (at least in our lab) seems to be ‘nybble’, to show its relation to ‘byte’, although this is falling into dissuse.

Word size is a complicated issue, especially when you start using 32- and 64- bit machines, where 64 bits can be refered to as either ‘DoubleWord’ (Motorola assembler) or ‘QuadWord’ (in Intel parlance)

A bit will be represented by multiple particles on the disk, probably thousands or millions. (But I don’t know a typical number offhand.)

Reading a bit means sensing the particles’ magnetic orientation, which can be done with a small circuit. A circular track of particles will be speeding past the read-write head as the disk spins. As this happens, electric current will briefly flow through the circuit in one direction or the other, depending on the orientation of the magnetic field passing by.

Writing a bit means setting the particles’ orientation of course — or more precisely the orientation of their collective magnetic field. (The particles themselves stay put, more or less.) This can be done by exposing them to sufficiently strong magnetic field, made by a small electromagnet, in the opposite direction from what’s desired.

Note that individual bits are never read or written to a disk in isolation. Disk drives are block-oriented devices; they always read and write entire blocks in one go. Typical block sizes range from one-half to four kilobytes.

Or “kilo-octets” I should say, to please our more orthodox readers.