| Kilooctets (ko) | Zettabits (Zbit) |
|---|---|
| 1 Kilooctet | 8 × 10-18 Zbit |
| 2 Kilooctets | 1.6 × 10-17 Zbit |
| 3 Kilooctets | 2.4 × 10-17 Zbit |
| 4 Kilooctets | 3.2 × 10-17 Zbit |
| 5 Kilooctets | 4 × 10-17 Zbit |
| 10 Kilooctets | 8 × 10-17 Zbit |
| 20 Kilooctets | 1.6 × 10-16 Zbit |
| 25 Kilooctets | 2 × 10-16 Zbit |
| 50 Kilooctets | 4 × 10-16 Zbit |
| 100 Kilooctets | 8 × 10-16 Zbit |
| Reference | Kilooctets (ko) | Zettabits (Zbit) |
|---|---|---|
| A plain text message (160 characters) | 0.16 ko | 1.28 × 10-18 Zbit |
| A three-minute MP3 | 3000 ko | 2.4 × 10-14 Zbit |
| A smartphone photo | 4000 ko | 3.2 × 10-14 Zbit |
| A high-definition film | 4000000 ko | 3.2 × 10-11 Zbit |
| A dual-layer Blu-ray disc | 50000000 ko | 4 × 10-10 Zbit |
The kilooctet is a unit of digital information equal to one thousand octets, and therefore to eight thousand bits. Its symbol is ko. It is the smallest of the everyday storage units, and for two decades it was the unit in which the whole capacity of a computer was described.
A kilooctet holds a thousand characters of unaccented text, which is about two hundred words, or a third of a page. The plain-text file of a short letter is a few kilooctets. Almost nothing else in modern computing is this small: an empty document from a word processor is already tens of kilooctets, because the file format carries formatting, fonts and metadata around the text.
The historical importance of the unit is hard to overstate. The Apple II shipped with 4 kilooctets of memory, the Commodore 64 was named for its 64, and the first IBM personal computer could address 640. Programs of real complexity — spreadsheets, word processors, games with graphics and sound — were written to fit inside those numbers, which required a discipline that has largely disappeared.
The kilooctet is also where the decimal and binary confusion began. Memory came in 1,024-octet units because addressing is binary, and everyone called that a kilooctet. Disc manufacturers counted in true thousands. The two conventions differ by 2.4 per cent, which was negligible at this scale, but the same error compounds at each step upward and reaches 10 per cent by the teraoctet.
The IEC resolved the ambiguity in 1998 by defining the kibioctet as 1,024 octets and leaving the kilooctet at exactly 1,000. Operating systems have adopted this unevenly: some report file sizes in true kilooctets, others still divide by 1,024 while writing ko, and a few now write Kio correctly.
Where the unit still appears daily is in network protocols and in the sizes of small resources on the web. A web page's stylesheet, an icon, a certificate, a configuration file, a database index page — all of these are measured in kilooctets, and the standard memory page on most processors is 4 kibioctets, close enough to 4 kilooctets for casual conversation but not for arithmetic.
One kilooctet equals 1,000 octets, 8,000 bits, 8 kilobits, or about 0.9766 kibioctets.
The zettabit is a unit of digital information equal to a thousand exabits, or a sextillion bits — a one followed by twenty-one zeros. Its symbol is Zbit. It is the largest unit in which real, measured quantities of data are still routinely reported, and it entered common use only in the last fifteen years.
A zettabit is 125 exaoctets. The phrase that made the prefix familiar was the zettabyte era, coined for the moment around 2016 when annual global internet traffic first exceeded one zettaoctet. That threshold arrived a little over a decade after annual traffic first passed one exaoctet, which gives a fair sense of the growth rate involved.
Estimates of the total quantity of data in existence are given in zettaoctets. The global datasphere — everything stored on every server, disc, phone and memory card, plus everything created and immediately discarded — is put at somewhere above a hundred zettaoctets, which is more than eight hundred zettabits. Nobody counts this directly; the figures come from manufacturing statistics and modelling.
Almost all of that is machine-generated and never read by anyone. Surveillance video that is recorded and overwritten, sensor telemetry, logs, backups and the many copies that redundancy demands account for the overwhelming majority. The portion that a human being has ever looked at is a tiny fraction of the total, and shrinking as a share every year.
The prefix zetta was adopted in 1991, along with yotta, at a conference that also had to consider what letters remained available. Z and Y were chosen partly because they were unused, and the pattern of naming from the Greek and Latin numerals had by then become loose. The 2022 additions of ronna and quetta continued the ladder upward for the same reason.
For perspective, a zettabit of information written as text in a single line of characters would stretch far beyond the solar system. The unit does not describe anything a person can hold or handle; it describes a civilisation's accumulated output, and it is useful precisely because that quantity now needs a name.
One zettabit equals 1,000 exabits, 125 exaoctets, or about 0.8470 zebibits.