Fact-checked by Grok 2 weeks ago

Kilobyte

A kilobyte (symbol: or ) is a of digital information used to measure capacity, file sizes, and in computing , traditionally equivalent to 1,024 bytes in the (base-2) numbering employed by computers. This value stems from the power-of-two structure of digital , where 1,024 equals 210, allowing efficient addressing in . Each byte itself consists of 8 bits, the fundamental digits (0 or 1) that represent . The term "kilobyte" originated in the mid-20th century as emerged, combining the SI prefix "kilo-" (meaning 1,000 or 103) with "byte," but early computer engineers adapted it to the convention of 1,024 bytes for practical reasons related to . By the and , it became standard for describing small media, such as 3.5-inch floppy disks holding up to 1.44 (1,440 kilobytes). However, this usage led to confusion with the strict definition, where one kilobyte equals exactly 1,000 bytes, particularly in contexts like hard drive capacities marketed by manufacturers. To clarify these distinctions, the (IEC) formalized binary prefixes in Amendment 2 to standard IEC 60027-2 in 1999, introducing the kibibyte (KiB) specifically for 1,024 bytes while reserving "kilobyte" for the decimal 1,000 bytes in non-binary contexts. Despite this standardization, the binary kilobyte (1,024 bytes) remains prevalent in software, operating systems, and technical documentation, though adoption of terms like KiB varies. Today, kilobytes measure modest data quantities, such as a plain-text (around 5 ) or a low-resolution image (100–800 KB).

Core Definitions

Decimal Kilobyte

The decimal kilobyte, abbreviated as , is defined as exactly 1,000 bytes, equivalent to $10^3 bytes, in alignment with the (SI) where the "kilo-" denotes a multiplication factor of 1,000. This definition stems from the system's decimal base, which facilitates straightforward scaling in scientific, , and general measurements by avoiding the irregularities of binary powers of two. The decimal kilobyte has been primarily adopted by international standards bodies, such as the (IEC), for specifying capacities in consumer-oriented applications like devices; for instance, hard drives marketed as 1 TB typically indicate $10^{12} bytes rather than a equivalent. In 1998, the IEC formalized this distinction by introducing binary-specific prefixes like kibibyte (KiB) to clarify usage, thereby establishing kB as explicitly decimal in standardized contexts. This approach contrasts with the traditional kilobyte used in some computing contexts.

Binary Kilobyte

The kilobyte is a of equal to exactly 1,024 bytes, equivalent to $2^{10} bytes. This definition arises from the nature of , where memory addressing and storage allocation are optimized using powers of two to simplify design and operations. Computers fundamentally operate in base-2, making multiples like 1,024 naturally aligned with components such as modules, which are organized in hierarchies, and systems that leverage power-of-two sizes for efficient . The term "kilobyte" has traditionally been used in to denote 1,024 bytes since the early . Mathematically, it is expressed as: $1 \, \mathrm{kB} = 2^{10} \, \mathrm{bytes} = 1{,}024 \, \mathrm{bytes} Larger units extend this pattern, for example: $1 \, \mathrm{MB} = 2^{20} \, \mathrm{bytes} = 1{,}048{,}576 \, \mathrm{bytes} This longstanding binary convention contributed to widespread confusion starting in the 1990s, as hard drive manufacturers began using the decimal kilobyte of 1,000 bytes for marketing larger apparent capacities, distinct from the SI alternative employed in some storage contexts. To address this ambiguity, the International Electrotechnical Commission introduced distinct binary prefixes in 1998, defining the kibibyte (KiB) as precisely 1,024 bytes while reserving kilobyte for decimal usage where applicable.

Historical Context

Origin and Early Adoption

The term "kilobyte" emerged in the mid-20th century as computing hardware advanced and the need arose to quantify larger units of digital information beyond the individual byte. The byte itself was coined in July 1956 by Werner Buchholz, an engineer working on the design of the Stretch computer (IBM 7030), to describe a fixed-size group of bits—typically eight—for and storage. The prefix "kilo-", borrowed from the to denote 1,000, was adapted in computing to the binary power of 2^10 = 1,024 due to addressing in powers of two, creating the initial convention for "K" as 1,024 in early machine specifications. Early adoption of the kilobyte concept appeared in the late 1950s and 1960s within mainframe documentation, where and storage capacities were described using "K bytes" or equivalent terms for 1,024 bytes. For instance, the family, announced in 1964, featured models with ranging from 8K bytes (8 × 1,024 bytes) to 512K bytes, reflecting the practical needs of binary-based systems for efficient addressing and scaling. storage, such as the 7-track tapes used with early IBM systems around 1956, began incorporating similar scaling for data volumes, though explicit "kilobyte" terminology solidified later in the decade. The first documented use of the full term "kilobyte" dates to 1970, combining "kilo-" with "byte" to describe 1,024 bytes in technical literature amid growing data handling requirements. Prior to formal definitions in the 1970s, usage showed ambiguity, with some references approximating 1,000 bytes in non-memory contexts like data transmission rates, while the binary 1,024 prevailed in core hardware descriptions due to its alignment with computer architecture. This early flexibility stemmed from the nascent state of digital units, evolving from decimal influences in pre-binary machines to binary standards in electronic computing.

Standardization Evolution

In the 1970s, standards organizations such as the (ANSI) and the Institute of Electrical and Electronics Engineers (IEEE) initially endorsed the binary definition of the kilobyte as 1024 bytes specifically for and contexts, reflecting the power-of-two prevalent in early digital systems. This approach was documented in key publications like the IEEE Standard Dictionary (IEEE Std 100, 1977 edition, revised 1979), which aligned with the practical needs of memory and storage addressing in -based hardware. Such endorsements helped formalize the binary kilobyte amid growing computational demands, though they did not yet address broader metric consistency. By the , ambiguities between and interpretations of the kilobyte led to widespread confusion, particularly in capacities where manufacturers increasingly adopted decimal prefixes (1000 bytes) while operating systems used ( bytes). This mismatch became evident in disputes, culminating in a 2003 class-action lawsuit filed against major computer manufacturers including Apple, , Gateway, , , , , and , alleging deceptive marketing of hard drive sizes—for instance, labeling a drive as "20 " when usable capacity appeared as only 18.6 under binary reporting. The case highlighted the need for clearer distinctions to avoid misleading users about actual storage availability. To resolve these conflicts, the (IEC) introduced a pivotal standard in December 1998 through Amendment 2 to IEC 60027-2 (published in 1999 as the second edition), formally defining the (kB) as exactly 1000 bytes in line with SI metric prefixes, while creating new binary prefixes like kibi (Ki) to denote 1024 bytes (thus, 1 KiB = 1024 bytes). This amendment aimed to eliminate ambiguity by reserving decimal prefixes for powers of 1000 and binary ones for powers of 1024, applying to and transmission fields. The change was driven by the need for precision as data volumes grew, ensuring compatibility with international metric standards. Subsequent standards from other bodies reflected varied adoptions. In the , the Joint Electron Device Engineering Council (), responsible for standards, retained the binary definition for memory modules like DDR , defining a kilobyte as bytes to match hardware addressing (e.g., in JESD209 series documents). Conversely, the National Institute of Standards and Technology (NIST) in the United States aligned with the IEC by endorsing decimal kilobytes (1000 bytes) for federal and general use, promoting the binary prefixes to avoid overlap. By 2008, the ISO/IEC 80000-13 standard reaffirmed the kilobyte as 1000 bytes, further solidifying global consensus on decimal usage for information quantities and influencing software implementations to adopt clearer labeling for storage devices. This evolution encouraged broader industry shifts toward decimal prefixes in marketing and reporting, reducing discrepancies in everyday applications.
YearMilestoneStandard BodyKey Definition
1956Informal adoption of binary kilobyte in computing literatureN/A (industry practice)1024 bytes (powers of 2)
1977/1979Endorsement of binary for computingIEEE (Std 100) / ANSI1024 bytes
1998/1999Introduction of binary prefixes; decimal for kBIEC (60027-2 Amendment 2)kB = 1000 bytes; KiB = 1024 bytes
2000sRetention of binary for semiconductors (e.g., JESD209)1024 bytes
2008Reaffirmation of decimal kilobyteISO/IEC (80000-13)kB = 1000 bytes

Modern Usage Contexts

In Data Storage

In data storage, manufacturers of hard disk drives (HDDs) and solid-state drives (SSDs) employ the decimal definition of the kilobyte—1 kB = 1,000 bytes—for labeling capacities, aligning with (SI) prefixes to simplify marketing. This convention results in advertised sizes that exceed those reported by operating systems, which frequently apply measurements (1 KiB = 1,024 bytes). For instance, a drive marketed as 1 TB holds 1,000,000,000,000 bytes, equivalent to roughly 931 GiB in terms, creating an apparent shortfall of about 7% due to the prefix discrepancy. Flash memory devices and USB drives follow a similar , with manufacturers using kilobytes for external labeling despite internal addressing in hardware designs. The Solid State Technology Association, which standardizes memory specifications, defines a kilobyte as 1,024 bytes (210) for capacities, reflecting architecture in chips like flash. However, consumer products such as USB flash drives are promoted using units (e.g., 1 GB = 1,000,000,000 bytes), leading to the same reporting differences when accessed via -oriented software. A representative example appears in electrically erasable programmable read-only memory () chips: a device rated at 1 kB capacity actually stores 1,024 bytes but is commonly marketed by its bit equivalent, such as 8 kilobits (8 kb = 8 × 1,024 bits = 8,192 bits = 1,024 bytes), emphasizing the foundation while using decimal-inspired for byte conversion. The evolution of standards, particularly the International Electrotechnical Commission's (IEC) 1998 adoption of distinct binary prefixes (e.g., KiB for 1,024 bytes), has influenced how operating systems display storage. macOS reports capacities in decimal terms to match manufacturer labels directly, showing unadjusted values like 1 TB as 1,000 GB. Conversely, utilities, such as the df command for filesystem usage, default to binary kilobytes (1 kB = 1,024 bytes) unless specified otherwise with options like --si for decimal scaling. As of 2025, providers vary in their approaches, but services like Amazon Simple Storage Service (S3) bill using gigabytes (1 GB = 1,073,741,824 bytes = 230 bytes), which mitigates some prefix confusion yet can still foster perceptions of 7-10% "lost" space among users accustomed to decimal expectations from .

In Network and Transmission

In network and transmission contexts, the kilobyte is predominantly defined using the decimal convention, where 1 kB equals 1000 bytes, to facilitate straightforward calculations aligned with units for rates. This approach simplifies and avoids the ambiguities associated with prefixes in dynamic transfer scenarios. metrics, such as kilobytes per second (kB/s), thus represent 1000 bytes per second, enabling clear reporting of throughput in protocols and devices. A representative example is Ethernet networking, where a 100 Mbps (megabits per second) link delivers approximately 12.5 MB/s (megabytes per second) of data, calculated by dividing the bit rate by 8 bits per byte and applying decimal multiples (100,000,000 bits/s ÷ 8 = 12,500,000 bytes/s = 12.5 × 10^6 bytes/s). This decimal usage ensures consistency in specifying link capacities across hardware standards. In protocols like TCP/IP, byte counts are handled as exact octets (8-bit units) within packet headers—for instance, the IP datagram's Total Length field measures up to 65,535 octets—but monitoring tools and reports convert these to decimal kilobytes for user-facing displays, such as HTTP response sizes in kB. This practice maintains precision in transmission while prioritizing readability in aggregated metrics. Historically, the ARPANET employed binary-based transmission, with network interfaces handling in binary streams independent of application layers, though line speeds like 50 kb/s used kilobit notations for throughput. In contrast, modern networks like adhere strictly to conventions; for example, a 1 Gbps (gigabits per second) throughput equates to approximately 125 MB/s, reflecting 1,000,000,000 bits/s ÷ 8 = 125,000,000 bytes/s. Internet service providers (ISPs) and streaming services further illustrate this dominance of decimal kilobytes. Netflix's bandwidth calculators estimate data usage in decimal gigabytes (GB = 10^9 bytes), such as up to 0.7 GB per hour for medium-quality streaming, to predict monthly consumption accurately without binary adjustments. Unlike static storage contexts, network transmission rarely adopts binary prefixes (e.g., KiB for 1024 bytes) to ensure alignment with SI decimal units, promoting global interoperability in protocol specifications and rate measurements.

Practical Examples and Comparisons

File Size Illustrations

A file of 1 (decimal, 1000 bytes) typically holds approximately 160-170 words in English, assuming an average word length of 5 characters plus a , though this can vary slightly with formatting and encoding. In the kilobyte definition ( bytes), the capacity increases marginally to about 170 words. For images, a small uncompressed of 100x100 , using 8 bits per pixel, occupies roughly 10 , excluding minor file header overhead. significantly reduces sizes for photographic images; for instance, a typical 1-megapixel photo that might be several megabytes uncompressed can be compressed to 50-200 while retaining acceptable quality, depending on the compression level and image complexity. In audio, one minute of uncompressed WAV audio at 8 kHz sampling rate in mono with 8-bit depth requires about 480 kB (decimal), suitable for low-fidelity applications like voice recordings. Using 16-bit depth, the same clip expands to approximately 960 kB. Practical examples include an average email without attachments, which ranges from 10-75 kB, primarily due to text content and basic HTML formatting. A short post on X (formerly Twitter) including a compressed image, within platform limits of 5 MB (mobile) or 15 MB (desktop) per media file, typically totals around 50 kB, with the image contributing the bulk after automatic optimization. To visualize kilobyte scales, an could compare 1 kB to everyday items, such as equating it to a short of about 150 words or a graphic.

Unit Conversion Examples

In and , converting between and kilobyte units ensures precise interpretation of and file sizes. The kilobyte (kB) represents exactly 1000 bytes, following the SI prefix "" for a of $10^3. In , the kilobyte, standardized as the kibibyte (KiB), equals 1024 bytes, or $2^{10}, to align with binary addressing in computer systems. Consequently, 1 kB (decimal) is approximately 0.9766 KiB, highlighting the roughly 2.4% difference that can accumulate in larger measurements. The fundamental conversion from decimal to binary units involves dividing the byte count by 1024. For example, to convert 1000 bytes (1 decimal kB), the calculation is $1000 / 1024 \approx 0.9766 KiB. This can be expressed generally as: \text{KiB} = \frac{\text{bytes (decimal)}}{1024} The reverse conversion, from binary to decimal, multiplies bytes by $1024 / 1000 or divides directly by 1000 after obtaining the byte equivalent. These operations are straightforward arithmetic but require attention to prefix conventions to avoid errors in applications like file synchronization. At larger scales, the divergence intensifies; for instance, 1 MB (), defined as 1,000,000 bytes, equates to approximately 976.5625 KiB ($1,000,000 / 1024). This discrepancy often explains variations in reported capacities between hardware manufacturers (typically ) and software displays (often ). Practical tools, such as IEC-compliant online binary prefix converters, simplify these calculations by automating the factor of 1024. For programmatic use, a basic function illustrates the process:
python
def decimal_to_kibibytes(bytes_decimal):
    return bytes_decimal / [1024](/page/1024)
Applying this, a 500 kB file in decimal terms contains 500,000 bytes, converting to $500,000 / [1024](/page/1024) \approx 488.281 KiB—useful for verifying sizes during cross-platform transfers where prefix interpretations differ.

References

  1. [1]
    What is a kilobyte (KB or Kbyte)? - TechTarget
    Feb 1, 2023 · A kilobyte (KB or Kbyte) is a unit of measurement for computer memory or data storage. Originally, a byte was considered a collection of 8 ...
  2. [2]
    What is a kilobyte? The unit of measurement explained - IONOS
    Aug 11, 2021 · The kilobyte (KB) is a standard unit of information that records the size of a file or data storage device.
  3. [3]
    Definitions of the SI units: The binary prefixes
    Once upon a time, computer professionals noticed that 210 was very nearly equal to 1000 and started using the SI prefix "kilo" to mean 1024. That worked well ...
  4. [4]
    Metric (SI) Prefixes | NIST
    Jan 13, 2010 · SI prefixes modify unit magnitude, using prefixes like deka (10), hecto (100), kilo (1000), deci (1/10), centi (1/100), and milli (1/1000).
  5. [5]
    About bits and bytes: prefixes for binary multiples - IEC
    This gives us the possibility to report the difference in a technically correct way: 1000 bytes are 1 kilobyte (Kbyte), and 1024 bytes are instead 1 kibibyte ( ...Missing: sources | Show results with:sources
  6. [6]
    Why are Bytes talked about in powers of 2? - Stack Overflow
    Sep 7, 2016 · The reason is that you do not only use bytes to store numbers, but also to address memory bytes that store numbers (or even other addresses).Missing: rationale | Show results with:rationale
  7. [7]
    Confusion with storage capacity (powers of 10 and 2) [duplicate]
    May 25, 2016 · The historical reason of using powers of 2 is that memory and hard disk are accessed by the CPU using an address space composed of lines on ...
  8. [8]
    Werner Buchholz Coins the Term "Byte", Deliberately Misspelled to ...
    In July 1956 German born American computer scientist Werner Buchholz Offsite Link coined the term byte Offsite Link as a unit of digital information.<|control11|><|separator|>
  9. [9]
    Gigabyte: Decimal vs. Binary - Coding Horror
    Sep 11, 2007 · Why propagate confusion? Instead of just learning “kilo- means 1000”, the rule is “kilo means 1000 when the base is meter, liter, gram, watt, ...
  10. [10]
    Kilobyte | computer science - Britannica
    It is measured as powers of two rather than powers of ten. When computer scientists first started talking in terms of large amounts of data, they just rounded ...
  11. [11]
    A look at IBM S/360 core memory: In the 1960s, 128 kilobytes ...
    Apr 15, 2019 · This article explains how core memory worked, how this core array was used in mainframes, and why core memory was so bulky.
  12. [12]
    Memory & Storage | Timeline of Computer History
    IBM 726 Magnetic tape. IBM 726 dual tape ... The outcome of the work was a read-only, 8-inch, 80 kilobyte floppy disk and disk drive—the world's first.
  13. [13]
    Kilobyte - Etymology, Origin & Meaning
    kilo- word-forming element meaning "one thousand," introduced in French 1795, when the metric system was officially adopted there.
  14. [14]
    Consumers sue over hard drive sizes | ZDNET
    Sep 19, 2003 · The result is that a hard drive described as being 20 gigabytes would actually have only 18.6 gigabytes of readable capacity, the lawsuit said.Missing: discrepancy date
  15. [15]
    [PDF] JESD209-2F - JEDEC STANDARD
    Hexadecimal number prefix. 0b. Binary number prefix. Byte. 8 bits. Word. 2 bytes = 16 bits. Kbit. 1024 bits. KByte. 1024 bytes. Mbit. 1,048,576 bits. MByte.
  16. [16]
    IEC 80000-13:2008 - Quantities and units — Part 13 - ISO
    IEC 80000-13:2008 provides names, symbols, and definitions for quantities and units in information science and technology, and is now withdrawn.Missing: kilobyte | Show results with:kilobyte
  17. [17]
    Why does my hard drive report less capacity than indicated on the ...
    Hard drive manufacturers market drives in terms of decimal (base 10) capacity. In decimal notation, one megabyte (MB) is equal to 1,000,000 bytes, one gigabyte ...
  18. [18]
    Why are hard drives never as large as advertised? - Super User
    Jul 15, 2009 · The technical reason is that the hard drive manufacturers sell you capacities in metric units. So a GB = 1,000,000,000 bytes by the metric ...Why do hard drives display a size lower than the true value?How to display byte sizes with binary prefixes? - Super UserMore results from superuser.com
  19. [19]
    [PDF] JESD220D.pdf - JEDEC STANDARD
    JEDEC standards and publications contain material that has been prepared, reviewed, and approved through the JEDEC Board of Directors level and subsequently ...<|separator|>
  20. [20]
    Available Capacity of the Drive is Smaller than the Drive Label
    Aug 20, 2018 · Systems will use both decimal and binary to measure the size. A drive's size is measured by using the total number of bytes that can be used on the drive.
  21. [21]
  22. [22]
    Why are EEPROM sizes measured in "k" or kbit and not kbyte or byte?
    Jan 4, 2014 · In both cases, the 1Kx4 (4kbit) chips are mapped into 1024 bytes of address space, but only the lower four data wires are connected (the extra ...
  23. [23]
    Serial EEPROM - Microchip Technology
    Microchip has been manufacturing serial EEPROMs for over 30 years and offers a full range of densities and communication protocols.
  24. [24]
    How storage capacity is measured on Apple devices
    Feb 26, 2025 · Storage capacity measured using the decimal system is slightly larger than storage capacity measured using the binary system (base 2). Binary ...Missing: IEC | Show results with:IEC
  25. [25]
    df(1) - Linux manual page - man7.org
    Otherwise, units default to 1024 bytes (or 512 if POSIXLY_CORRECT is set). The SIZE argument is an integer and optional unit (example: 10K is 10*1024). Units ...
  26. [26]
    Amazon S3 FAQs - Cloud Object Storage - AWS
    5,257,039,970,304 Byte-Hours x (1 GB / 1,073,741,824 bytes) x (1 month / 744 hours) = 6.581 GB-Month The cost is calculated based on the current rates for your ...
  27. [27]
    Bits, Bytes and Bandwidth Reference Guide - SpeedGuide
    Kilobits per second is usually shortened to kbps or Kbps**. Some sources define a kilobit to mean 1,024 bits. Although the bit is a unit of the binary number ...Missing: convention | Show results with:convention
  28. [28]
    None
    ### Summary: TCP/IP Byte Counts in RFC 791
  29. [29]
    [PDF] Remembering the BBN ARPANET Project - Walden Family
    System design (ARPANET characteristics) a. Reliable transmission b. Network transmitted binary data (application independent) c. Dynamic routing d. In-band ...
  30. [30]
    Bandwidth conversion - toolstud.io
    How do you convert from one into the other? This bandwidth calculator tool will help you convert from one to the other. Speed.Missing: convention | Show results with:convention
  31. [31]
    How to control how much data Netflix uses
    Control data usage: from a web browser go to your account page, then profile and parental controls. From a mobile device go to My Netflix and select ...Missing: kilobytes decimal
  32. [32]
    If I would have a 1GB text file (txt), approximately how many ... - Quora
    Oct 9, 2019 · The average word is 5 letters long. 1kB can hold 167 words. So a GB would be way over 167 million words. Also, an A4 page can have 500 words approximately.How many words are in 1kb? - QuoraWhat is the average size of a short message service (SMS) in bytes ...More results from www.quora.com
  33. [33]
    File Size Calculations - 101 Computing
    Feb 24, 2021 · Text file size is calculated by bits per character x characters. Picture size is color depth x width x height. Sound size is sample rate x ...
  34. [34]
    Image File Size Calculator
    This image file size calculator will help you estimate the file size of an uncompressed image file at a given pixel count and bit depth.
  35. [35]
    What is the "optimal" file size of JPEG images with respect to their ...
    Jan 9, 2019 · On average, JPEG's sweet spot is around one bit per pixel. This will of course vary depending on image content.JPG and file sizes - BASIC facts for dummyDoes simply opening and closing a JPEG file decrease image quality?More results from photo.stackexchange.com
  36. [36]
    Audio File Size Calculator
    This audio file size calculator will help you estimate how much space an uncompressed audio file will take up on your computer's storage.
  37. [37]
    Audio File Size Calculator - Colin Crawley
    This utility calculates the size of audio files (both uncompressed, PCM/IEEE FP audio, such as .WAV/ .W64/ .RF64, .AIFF/.AIF and also lossy compressed files)
  38. [38]
    How many email users are there in 2025? (and other email usage ...
    Apr 8, 2024 · An average email without attachments is 75KB. Studies show that an ... size, not including any attachments. (Source: MailMeteor).
  39. [39]
    X (Twitter) Image Sizes - Metricool
    Ideal image size: 400×400 pixels · Minimum image size: 200×200 pixels · Aspect ratio: 1:1 · Maximum file size: 2MB · Supported file formats: JPEG, GIF, PNG.(Twitter) Profile Picture Size · (Twitter) Header Size · (Twitter) In-Stream Image Size<|control11|><|separator|>
  40. [40]
    Understanding file sizes | GreenNet
    Table of approximate file sizes ; 1,000, 1 kB, Half a page of unformatted text; a very short email; an icon or small button image ; 8,000, 8 kB, Typical size of ...