Explaining SI Unit for Computer Storage

Most of us who use computers either in our personal or professional lives will be familiar with the idea of ​​computer storage and some of the units which are used to design storage capacity at the lower end of the spectrum. The following article provides an overview of all of the units that have, as of present, been classified as SI units (International System of Units) including the familiar kilo-, mega- and gigabytes.

Before looking at its multiples it is worth becoming familiar with the concept of a byte (B) and what it represents. The fundamental unit of computer information storage is actually a bit (short for a binary digit) which is a binary concept in that it exists in one of two states, 1 or 0. A byte in turn is derived from eight bits because that was generally the number needed to form the smallest unit of meaningful data – a single character of text. Therefore, a byte is usually now equivalent to an 'octet' of bits although it can and has varied depending on the hardware in question. The word itself actually originated in 1956 as an intentionally incorrect spelling of the word bite to avoid corruption back to its smaller constituent, the bit.

Bytes can be measured in decimal (multiples of ten) or binary (multiples of two), again depending on the manufacturer of the hardware in question and unofficial convention. The following are decimal multiples as specified by the SI classification but the binary use of the terms is actually far more prevalent in practice. For every 1,000 bytes, the nearest binary equivalent would be 1,024 bytes. As binary systems us the concept of two states (eg, true / false, off / on, 1/0, positive / negative) to represent information the binary multiples use doubling rather than factors of ten. To achieve the nearest equivalent value to 1,000 before the sequence 2, 4, 8, 16, 32, 64, 128, 256, 512, 1,024 is used; 1,024 being two to the power of ten (ie, 210). At the lower end of the scale, binary and decimal equivalents are fairly close to each other (1,000 vs 1,024 is different by 2.4%), but as the scales grow the difference is magnified to the greater that the nearest equivalent multiples are almost 21% apart when considering 280 and 1024.

To avoid confusion between decimal and binary values, binary specific units have been standardized by the International Electrotechnical Commission (IEC) although the SI terms are still most frequently used to refer to the binary values.

Kilobyte

• kB (KB for binary)
• Number of bytes:
  • Decimal – 103
  • Binary – 210
  • Difference between decimal and binary figures – 2.40%
• IEC equivalent:
  • Prefix – Kibi-
  • Symbol – KiB
• Example Storage: A text file of approximately 150 words (ie, approx 1,000 characters)
• First Used: "kilo-" 1940s; "kilobyte" circa mid 1960s

Megabyte

• MB
• Number of bytes:
  • Dec. – 106
  • Bin. – 220
  • Diff. – 4.86%
• IEC:
  • Mebi-
  • MiB
• Example Storage:
  • A text file / book of approximately 500 pages of 2,000 characters per page.
  • (If 8 bits per pixel, ie, 256 color depth) A 1,024×1,024 resolution image
• First Used: "mega-" 1940s

Gigabyte

• GB
• Number of bytes:
  • Dec. – 109
  • Bin. – 230
  • Diff. – 7.37%
• IEC:
  • Gibi-
  • GiB
• Example Storage: One hour of standard definition video
• First Used: "giga-" defined in 1960

Terabyte

• TB
• Number of bytes:
  • Dec. – 1012
  • Bin. – 240
  • Diff. – 9.95%
• IEC:
  • Tebi-
  • TiB
• Example Storage: Around 17,000 hours of compressed audio
• First Used: "terra-" defined in 1960

Petabyte

• PB
• Number of bytes:
  • Dec. – 1015
  • Bin. – 250
  • Diff. – 12.59%
• IEC:
  • Pebi-
  • PiB
• Example Storage: The entire video storage of Netflix
• First Used: "peta-" defined in 1975

Exabyte

• EB
• Number of bytes:
  • Dec. – 1018
  • Bin. – 260
  • Diff. – 15.29%
• IEC:
  • Exbi-
  • EiB
• Example Storage:
  • Estimated internet traffic per day in 2010
  • Estimated cloud hosting capacity of Amazon
• First Used: "exa-" defined in 1975

Zettabyte

• ZB
• Number of bytes:
  • Dec. – 1021
  • Bin. – 270
  • Diff. – 18.06%
• IEC:
  • Zebi-
  • ZiB
• Example Storage: All human speech ever detected could be stored as an estimated 42ZB of compressed audio
• First Used: "zetta-" defined in 1991

Yottabyte

• YB
• Number of bytes:
  • Dec. – 1024
  • Bin. – 280
  • Diff. – 20.89%
• IEC:
  • Yobi-
  • YiB
• Example Storage:
  • All human speech in approx. 0.042 YB
  • By some estimates the entire digital data stored on earth
• First Used: "yotta-" defined in 1991