[comp.arch] standard size names

jk3k+@andrew.cmu.edu (Joe Keane) (03/02/88)

We all know a `nibble' is 4 bits, and a `byte' is 8.  But what are 16, 32 and 
64 bits?  In C terminology you've got `short', `long' and `long long' (yes i 
know they could be different).  In VAX terminology they're `word', `longword' 
and `quadword'.  Most 32-bit computers use `halfword', `word' and `double 
word'.  So does anyone have a decent set of names for these things?

--Joe

fouts@orville.nas.nasa.gov (Marty Fouts) (03/03/88)

In article <UW=1uvy00XoH4DZktq@andrew.cmu.edu> jk3k+@andrew.cmu.edu (Joe Keane) writes:
>
>We all know a `nibble' is 4 bits, and a `byte' is 8.  But what are 16, 32 and 
>64 bits?  In C terminology you've got `short', `long' and `long long' (yes i 
>know they could be different).  In VAX terminology they're `word', `longword' 
>and `quadword'.  Most 32-bit computers use `halfword', `word' and `double 
>word'.  So does anyone have a decent set of names for these things?
>
>--Joe

I like

4	nibble
8	byte
16	mouthfull
32	gulp
64	snack
80	meal (IEEE)
128	pigout

myself.

Marty

jlg@a.UUCP (Jim Giles) (03/03/88)

> 

On the Cray, these are called parcel (16-bit), halfword (32-bit), and
word (64-bit).  I like this terminology since it implies that machines
with smaller word sizes are just toys :).  Of course, there's never been
complete agreement on those other terms anyway (CDC machines all used to 
have 6-bit characters - everyone called them bytes).  

My own feeling is that 'word' should always refer to the resolution of
memory and/or disk addresses.  Either that, or we need another term to
describe such a unit.  Of course, this would mean that VAXen and such
ilk would have 8-bit words :).

J. Giles
Los Alamos

davidsen@steinmetz.steinmetz.UUCP (William E. Davidsen Jr) (03/03/88)

In article <UW=1uvy00XoH4DZktq@andrew.cmu.edu> jk3k+@andrew.cmu.edu (Joe Keane) writes:
| We all know a `nibble' is 4 bits, and a `byte' is 8.  But what are 16, 32 and 
| 64 bits?  In C terminology you've got `short', `long' and `long long' (yes i 

Actually you forgot going toward zero and names. I have heard the term
'tayste' used for half a nybble, and everyone agrees that a bit is a
bit.

Personally I think half a nybble should be a 'quarter'. Everyone knows
that two bits is a quarter (Sorry, couldn't resist).
-- 
	bill davidsen		(wedu@ge-crd.arpa)
  {uunet | philabs | seismo}!steinmetz!crdos1!davidsen
"Stupidity, like virtue, is its own reward" -me

abh@POGO.CAMELOT.CS.CMU.EDU (Andrew Hastings) (03/04/88)

In article <499@a.UUCP>, jlg@a.UUCP (Jim Giles) writes:
>						 (CDC machines all used to 
> have 6-bit characters - everyone called them bytes).  
Well, not quite.  A CDC `byte' was 12 bits, or 2 6-bit characters, or 1 PP
word, or 1/5 CM word.

-Andrew Hastings		abh@cs.cmu.edu			412/268-8734

devine@cookie.dec.com (Bob Devine) (03/04/88)

> We all know a `nibble' is 4 bits, and a `byte' is 8.
> But what are 16, 32 and 64 bits?

  This is something that's come up at least once a year
since I've been reading usenet postings.

  My suggestion for 64 bits has been to honor of Seymour Cray
because he has been at the forefront of 64 bit design for
about two decades.  64 bits == a "crayte".

Bob Devine  ...!decwrl!cookie.dec.com!devine

ccplumb@watmath.waterloo.edu (Colin Plumb) (03/05/88)

davidsen@crdos1.UUCP (bill davidsen) wrote:
>jk3k+@andrew.cmu.edu (Joe Keane) writes:
>> We all know a `nibble' is 4 bits, and a `byte' is 8.  But what are 16, 32
>> and 64 bits?  In C terminology you've got `short', `long' and `long long'
>
>Actually you forgot going toward zero and names. I have heard the term
>'tayste' used for half a nybble, and everyone agrees that a bit is a
>bit.
>
>Personally I think half a nybble should be a 'quarter'. Everyone knows
>that two bits is a quarter (Sorry, couldn't resist).

Helios, the promised OS for Atari's promised Transputer box, used the term
"morsel" for 2 bits.  A quote from the spec:

"I could not recall any name for a 2-bit field, but my thesaurus lists
bite, nibble, and morsel in that order, so the choice was obvious."

Two bits is a quarter?  You're *still* using pieces of eight down there? :-)
--
	-Colin (watmath!ccplumb)

Zippy says:
I'm in direct contact with many advanced fun CONCEPTS.

haynes@ucscc.UCSC.EDU (99700000) (03/08/88)

Of course in Olivetti machines, the Italians being noted for their musical
abilities, four bits is a quartet and eight bits is an octet.
haynes@ucscc.ucsc.edu
haynes@ucscc.bitnet
..ucbvax!ucscc!haynes

lisper@yale.UUCP (Bjorn Lisper) (03/09/88)

In article <2246@saturn.ucsc.edu> haynes@ucscc.UCSC.EDU (Jim Haynes) writes:
>Of course in Olivetti machines, the Italians being noted for their musical
>abilities, four bits is a quartet and eight bits is an octet.

This has actually been proposed in Swedish too. It is certainly more precise
than talking about bytes and words which means different things on different
machines.

Bjorn Lisper

aglew@ccvaxa.UUCP (03/10/88)

..> quartet, octet...

One of my pet peeves:

1 bit, singlet, 1b
2 bits, doublet, 2b
3 bits, triplet, 3b
4 bits, quadlet/quartet, (nibble), 4b
8 bits, octlet/octet, (byte), 8b, 1B
16 bits, hexlet/hextet, (word/halfword/whatever...), 16b, 2B
32 bits, triclet, (word/doubleword/whatever), 32b, 4B
64 bits, sexlet, (word,doubleword,longword,quadword), 64b, 8B

Discourage the terms in parentheses.



Andy "Krazy" Glew. Gould CSD-Urbana.    1101 E. University, Urbana, IL 61801   
    aglew@gould.com     	- preferred, if you have MX records
    aglew@xenurus.gould.com     - if you don't
    ...!ihnp4!uiucuxc!ccvaxa!aglew  - paths may still be the only way
   
My opinions are my own, and are not the opinions of my employer, or any
other organisation. I indicate my company only so that the reader may
account for any possible bias I may have towards our products.

peter@athena.mit.edu (Peter J Desnoyers) (03/11/88)

In the standards world:

   1 bit is a bit
   8 bits is an octet (god help us, not an 'octlet')
   and that's unfortunately all.

Of course, in information theory:

   1 nat = 2.718282... bits    (the 'natural' digit, instead of 'binary'
				digit - take the natural log when 
				calculating information instead of log2)

but this isn't a useful word length for computers. (You don't even see
it in many information theory papers)

Anyway, _please_ don't call 8 bits an `octlet'. It already has a name.


				Peter Desnoyers