Brain Dump

Byte

Tags
comp-sci

Is a unit of digital information consisting of 8-bits.

The size of a byte has in some cases been ambiguous, in which case the term octet represents exactly 8-bits.

Bytes are commonly prefixed with ISO standard values when representing large values. This follows the standard kilo, mega, giga metric prefix as shown in tbl:kilo. An alternative scheme that more closely relates to the binary nature of hard disk drives is the kibi, mibi, gibi prefix standard as shown in tbl:kibi. Confusingly in the real world the conventions vary. When a file is displayed in the OS a KB is treated the same as a KiB, however when talking about computer networks, CDs or other storage KB is not the same as KiB and is the ISO metric definition described above.

This is a historical quirk was brought by a clash between network developers and memory/hard storage developers. Hard storage and memory developers found that if a bit could take one of two states, it would be natural to call a Kilo- prefix 1024 because it was about 1000. Network developers had to deal with bits, real-time signal processing, and various other factors, so they went with the already accepted convention that Kilo- means 1000 of something.

Table 1: Values of byes with the standard ISO metric prefix scheme. tbl:kilo
PrefixByte Value
KB\( 10^3 \) B
MB\( 10^6 \) B
GB\( 10^9 \) B
Table 2: Values of byes with the base-2 kibi-byte prefix scheme. tbl:kibi
PrefixByte Value
KiB\( 2^{10} \) B
MiB\( 2^{20} \) B
GiB\( 2^{30} \) B

Links to this note