log in or sign up to track your course progress, gain accessibility to final exams, and get a free certificate that completion!

"In computing, a word is the herbal unit of data offered by a specific processor design. A indigenous is a fixed-sized item of data handled as a unit through the instruction set or the hardware that the processor. The number of bits in a word (the word size, indigenous width, or indigenous length) is an essential characteristic the any details processor architecture or computer architecture."

Incomputing, awordis the organic unit of data provided by a particularprocessordesign. A native is a fixed-sizedpiece of datahandled together a unit by theinstruction setor the hardware the the processor. The number ofbitsin a native (theword size,word width, orword length) is critical characteristic of any specific processor architecture orcomputer architecture.

You are watching: How long is a word in bytes

The size of a indigenous is reflect in many elements of a computer"s structure and also operation; the majority of theregistersin a processor are usually native sized and also the biggest piece the data that can be transferred to and also from theworking memoryin a single operation is a native in many (not all) architectures. The biggest possibleaddresssize, offered to designate a ar in memory, is typically a hardware indigenous (here, "hardware word" way the full-sized herbal word the the processor, as opposed to any kind of other definition used).

Modern processors, includingembedded systems, usually have a word dimension of 8, 16, 24, 32, or 64bits, while modern-day general purpose computer systems usually use 32 or 64 bits. Special purpose digital processors, together asDSPsfor instance, may use various other sizes, and also many other sizes have been used historically, including 9,12,18, 24,26,36,39, 40, 48, and 60 bits. Theslabis an instance of a mechanism with an previously word size. Several of the earliest computers (and a few modern together well) usedBCDrather 보다 plainbinary, commonly having a word dimension of 10 or 12 decimal digits, and also some earlydecimal computershad no addressed word size at all.

The size of a word have the right to sometimes differ from the expected due tobackward compatibilitywith previously computers. If many compatible sports or a family members of processors re-publishing a usual architecture and instruction collection but different in their word sizes, their documentation and also software may become notationally complicated to accommodate the difference (seeSize familiesbelow).

Uses the words

Depending on how a computer is organized, word-size units might be used for:

Fixed suggest numbersHolders forfixed point, usuallyinteger, number values might be accessible in one or in several different sizes, however one of the sizes accessible will practically always be the word. The various other sizes, if any, are most likely to it is in multiples or fountain of the word size. The smaller sizes are usually used just for effective use the memory; once loaded right into the processor, their worths usually get in a larger, word sized holder.Floating suggest numbersHolders forfloating pointnumerical values are commonly either a indigenous or a lot of of a word.AddressesHolders for memory addresses should be the a size capable of to express the needed selection of values however not it is in excessively large, so regularly the size offered is the word though it can additionally be a multiple or fraction of words size.RegistersProcessor registersare designed through a size suitable for the kind of data castle hold, e.g. Integers, floating suggest numbers or addresses. Many computer system architectures usage "general purpose registers" that deserve to hold any of several species of data, these registers must be sized to hold the biggest of the types, historically this is the word dimension of the architecture though progressively special purpose, larger, registers have been included to attend to newer types.Memory-processor transferWhen the processor reads indigenous the memory subsystem into a register or writes a register"s value to memory, the amount of data moved is regularly a word. Historically, this lot of bits which can be transferred in one bike was likewise called acatenain some environments (such together theBullGAMMA 60(fr)).In straightforward memory subsystems, words is transferred over the memorydata bus, which typically has a width of a word or half-word. In storage subsystems that usecaches, the word-sized transport is the one in between the processor and the an initial level of cache; at reduced levels that thememory hierarchylarger move (which space a lot of of words size) are usually used.Unit of resolve resolutionIn a offered architecture, successive address values designate successive units the memory; this unit is the unit of deal with resolution. In most computers, the unit is either a character (e.g. A byte) or a word. (A couple of computers have actually used bit resolution.) If the unit is a word, climate a larger amount of memory deserve to be accessed utilizing an address of a given size in ~ the cost of added complexity to access individual characters. ~ above the other hand, if the unit is a byte, then individual characters can be addressed (i.e. Selected during the storage operation).InstructionsMachine instructionsare usually the dimension of the architecture"s word, such together inRISC architectures, or a multiple of the "char" dimension that is a portion of it. This is a natural selection since instructions and also data generally share the exact same memory subsystem. InHarvard architecturesthe word sizes of instructions and data need not be related, as instructions and also data space stored in various memories; for example,the processor in the 1ESS digital telephone switchhad 37-bit instructions and also 23-bit data words.

Word size choice

When a computer architecture is designed, the an option of a word size is of comprehensive importance. There are style considerations i beg your pardon encourage details bit-group sizes for particular uses (e.g. For addresses), and also these considerations suggest to different sizes for different uses. However, considerations of economic climate in design strongly push for one size, or a very few sizes connected by multiples or fountain (submultiples) come a primary size. That desired size i do not care the word size of the architecture.

Charactersize remained in the previous (pre-variable-sizedcharacter encoding) one of the influences on unit of deal with resolution and the an option of indigenous size. Prior to the mid-1960s, personalities were most often stored in six bits; this permitted no much more than 64 characters, for this reason the alphabet was limited to top case. Because it is effective in time and space to have the word size be a lot of of the personality size, word size in this duration were normally multiples the 6 bits (in binary machines). A common an option then to be the36-bit word, i beg your pardon is also a an excellent size because that the numeric nature of a floating point format.

After the advent of theIBMSystem/360design, which provided eight-bit characters and also supported lower-case letters, the standard dimension of a personality (or more accurately, abyte) became eight bits. Indigenous sizes after that were naturally multiples the eight bits, through 16, 32, and also 64 bits being frequently used.

Variable word architectures

Early an equipment designs had some that used what is frequently termed avariable indigenous length. In this kind of organization, a numeric operand had actually no addressed length however rather its finish was detected as soon as a character through a one-of-a-kind marking, frequently calledword mark, to be encountered. Such machines frequently usedbinary coded decimalfor numbers. This class of machines contained theIBM 702,IBM 705,IBM 7080,IBM 7010,UNIVAC 1050,IBM 1401, andIBM 1620.

Most that these makers work ~ above one unit of storage at a time and since every instruction or datum is number of units long, each instruction takes number of cycles just to accessibility memory. These equipments are often quite slow due to the fact that of this. For example, instruction fetches ~ above anIBM 1620 model Itake 8 cycles simply to read the 12 digits of the instruction (theModel IIreduced this to 6 cycles, or 4 cycles if the instruction did not need both resolve fields). Accuse execution take it a fully variable variety of cycles, depending upon the size of the operands.

Word and byte addressing

The memory model of an architecture is strongly influenced by words size. In particular, the resolution the a memory address, that is, the the smallest unit that have the right to be designated by an address, has regularly been liked to be the word. In this approach, deal with values which different by one designate adjacent memory words. This is organic in makers which deal nearly always in native (or multiple-word) units, and has the benefit of enabling instructions to usage minimally sized areas to save addresses, which deserve to permit a smaller sized instruction size or a larger range of instructions.

When byte processing is to it is in a far-reaching part the the workload, the is usually an ext advantageous to usage thebyte, fairly than the word, together the unit of address resolution. This enables an arbitrarily character in ~ a character string to be handle straightforwardly. A word have the right to still it is in addressed, yet the address to be offered requires a couple of more bits than the word-resolution alternative. Words size requirements to be an creature multiple the the character dimension in this organization. This addressing approach was offered in the IBM 360, and also has to be the many common approach in equipments designed since then.

Individual bytes have the right to be accessed ~ above a word-oriented device in one of two ways. Bytes have the right to be manipulated by a combination of shift and mask work in registers. Moving a solitary byte native one arbitrary location to another may require the tantamount of the following:

LOAD words containing the resource byteSHIFTthe source word come align the desired byte to the correct place in the target wordANDthe resource word through a mask to zero out all but the desired bitsLOAD the word containing the target byteAND the target word v a mask to zero the end the target byteORthe registers containing the resource and target words come insert the source byteSTORE the result back in the target location

Alternatively plenty of word-oriented devices implement byte operations through instructions using specialbyte pointersin registers or memory. For an example thePDP-10byte pointer included the dimension of the byte in bits (allowing different-sized bytes to it is in accessed), the bit position of the byte within the word, and the word resolve of the data. Instructions can automatically adjust the tip to the next byte on, for example, load and deposit (store) operations.

Powers of two

Different quantities of storage are used to store data values v different levels of precision. The commonly used sizes space usually apower of twomultiple of the unit of deal with resolution (byte or word). Converting the table of contents of things in range into the resolve of the item climate requires only ashiftoperation rather than a multiplication. In some cases this connection can additionally avoid the usage of department operations. As a result, most contemporary computer designs have actually word sizes (and other operand sizes) that are a strength of two times the size of a byte.

Size families

As computer designs have grown more complex, the main importance that a single word size to an architecture has decreased. Although much more capable hardware can use a more comprehensive variety of size of data, market pressures exert pressure to maintainbackward compatibilitywhile extending processor capability. Together a result, what can have to be the main word size in a fresh style has to coexist as an alternate size come the initial word size in a behind compatible design. The original word size remains accessible in future designs, forming the communication of a dimension family.

In the mid-1970s,DECdesigned theVAXto it is in a successor of thePDP-11. Castle usedwordfor a 16-bit quantity, whilelongwordreferred to a 32-bit quantity. This to be in contrast to previously machines, wherein the organic unit of addressing memory would be referred to as aword, if a quantity that is one half a word would certainly be called ahalfword. In fitting with this scheme, a VAXquadwordis 64 bits.

Another example is thex86family, of i m sorry processors that three different word lengths (16-bit, later 32- and also 64-bit) have been released. As software is routinelyportedfrom one word-length to the next, someAPIsand documentation define or refer to an older (and for this reason shorter) word-length than the full word size on the CPU that software may be compiled for. Also, comparable to how bytes are offered for small numbers in plenty of programs, a much shorter word (16 or 32 bits) might be offered in contexts wherein the range of a more comprehensive word is not needed (especially where this deserve to save substantial stack an are or cache storage space). Because that example, Microsoft"sWindows APImaintains theprogramming languagedefinition ofWORDas 16 bits, regardless of the fact that the API may be supplied on a 32- or 64-bit x86 processor, wherein the standard word size would it is in 32 or 64 bits, respectively. Data frameworks containing such different sized words describe them asWORD(16 bits/2 bytes),DWORD(32 bits/4 bytes) andQWORD(64 bits/8 bytes) respectively. A similar phenomenon has developed inIntel"sx86assembly language– since of the support for assorted sizes (and backward compatibility) in the indict set, some instruction mnemonics bring "d" or "q" identifiers denoting "double-", "quad-" or "double-quad-", which space in terms of the architecture"s initial 16-bit native size.

In general, brand-new processors have to use the very same data native lengths and also virtual resolve widths as an larger processor come havebinary compatibilitywith that older processor.

Often very closely written source code – written withsource code compatibilityandsoftware portabilityin psychic – can be recompiled to operation on a variety of processors, even ones with various data word lengths or different attend to widths or both.

See more: How Many Pictures On A 16Gb Card, How Much Does A 16Gb Sandisk Memory Card Hold


This job-related is license is granted under a an imaginative Commons Attribution-ShareAlike 3.0 License.