How many bits is a "word"?

Question

This is from the book Assembly Language Step By Step, Jeff Duntemann:

Here’s the quick tour: A bit is a single binary digit, 0 or 1. A byte is 8 bits side by side. A word is 2 bytes side by side. A double word is 2 words side by side. A quad word is 2 double words side by side.

And this is from the book Principles of Computer Organization and Assembly Language: Using the Java Virtual Machine, Patrick Juola:

For convenience, 8 bits are usually grouped into a single block, conventionally called a byte. The next-largest named block of bits is a word. The definition and size of a word are not absolute, but vary from computer to computer. A word is the size of the most convenient block of data for the computer to deal with.

So is a word 2 bytes (16 bits), or is it the most convenient block of data for the computer to deal with? (I am also not sure what this means..)

Answer 1

I'm not familiar with either of these books, but the second is closer to current reality. The first may be discussing a specific processor.

Processors have been made with quite a variety of word sizes, not always a multiple of 8.

The 8086 and 8087 processors used 16 bit words, and it's likely this is the machine the first author was writing about.

More recent processors commonly use 32 or 64 bit words.

In the 50's and 60's there were machines with words sizes that seem quite strange to us now, such as 4, 9 and 36. Since about the 70's word size has commonly been a power of 2 and a multiple of 8.

Answer 2

On x86/x64 processors, a byte is 8 bits, and there are 256 possible binary states in 8 bits, 0 thru 255. This is how the OS translates your keyboard key strokes into letters on the screen. When you press the 'A' key, the keyboard sends a binary signal equal to the number 97 to the computer, and the computer prints a lowercase 'a' on the screen. You can confirm this in any Windows text editing software by holding an ALT key, typing 97 on the NUMPAD, then releasing the ALT key. If you replace '97' with any number from 0 to 255, you will see the character associated with that number on the system's character code page printed on the screen.

If a character is 8 bits, or 1 byte, then a WORD must be at least 2 characters, so 16 bits or 2 bytes. Traditionally, you might think of a word as a varying number of characters, but in a computer, everything that is calculable is based on static rules. Besides, a computer doesn't know what letters and symbols are, it only knows how to count numbers. So, in computer language, if a WORD is equal to 2 characters, then a double-word, or DWORD, is 2 WORDs, which is the same as 4 characters or bytes, which is equal to 32 bits. Furthermore, a quad-word, or QWORD, is 2 DWORDs, same as 4 WORDs, 8 characters, or 64 bits.

Note that these terms are limited in function to the Windows API for developers, but may appear in other circumstances (eg. the Linux dd command uses numerical suffixes to compound byte and block sizes, where c is 1 byte and w is bytes).

Answer 3

The second quote is correct, the size of a word varies from computer to computer. The ARM NEON architecture is an example of an architecture with 32-bit words, where 64-bit quantities are referred to as "doublewords" and 128-bit quantities are referred to as "quadwords":

A NEON operand can be a vector or a scalar. A NEON vector can be a 64-bit doubleword vector or a 128-bit quadword vector.

Normally speaking, 16-bit words are only found on 16-bit systems, like the Amiga 500.

Answer 4

This is from the book Hackers: Heroes of the Computer Revolution by Steven Levy.

.. the memory had been reduced to 4096 "words" of eighteen bits each. (A "bit" is a binary digit, either a 1 or 0. A series of binary numbers is called a "word").

As the other answers suggest, a "word" does not seem to have a fixed length.

Answer 5

In addition to the other answers, a further example of the variability of word size (from one system to the next) is in the paper Smashing The Stack For Fun And Profit by Aleph One:

We must remember that memory can only be addressed in multiples of the word size. A word in our case is 4 bytes, or 32 bits. So our 5 byte buffer is really going to take 8 bytes (2 words) of memory, and our 10 byte buffer is going to take 12 bytes (3 words) of memory.

Answer 6

"most convenient block of data" probably refers to the width (in bits) of the WORD, in correspondance to the system bus width, or whatever underlying "bandwidth" is available. On a 16 bit system, with WORD being defined as 16 bits wide, moving data around in chunks the size of a WORD will be the most efficient way. (On hardware or "system" level.)

With Java being more or less platform independant, it just defines a "WORD" as the next size from a "BYTE", meaning "full bandwidth". I guess any platform that's able to run Java will use 32 bits for a WORD.

Answer 7

Another instance of a book citing the variable length of the Word is Operating System Concepts by Sileberschatz, Galvin, Gagne where the authors in Chapter 1 page 6 state:

A less common term is "word", which is a given computer architecture's native storage unit. A word is generally made up of one or more bytes. For example, a computer may have instructions to move 64-bit (8-byte) words.

Edit
Yet another instance of a book citing I found today is "C Primer Plus 6th Edition" by Stephen Prata, where in Chapter 3 page 60, he writes:

A word is the natural unit of memory for a given computer design. For 8-bit microcomputers, such as the original Apples, a word is just 8 bits. Since then, personal computers moved up to 16-bit words, 32-bit words, and, at the present, 64-bit words. Larger word sizes enable faster transfer of data and allow more memory to be accessed.