bader@b.psy.cmu.edu (Miles Bader) (12/16/85)
Well, the 8088 *IS* a 16 bit processor internally-- i.e., 16 bit data paths, 16 bit alu. But the 68000 is ALSO a 16 bit processor internally-- 16 bit data path and alu. It just happens to have operations on longs built in... Not that I like the 8088, of course. [processor wars!] -Miles
dml@loral.UUCP (Dave Lewis) (12/21/85)
In article <730@brl-tgr.ARPA> bader@b.psy.cmu.edu (Miles Bader) writes: >Well, the 8088 *IS* a 16 bit processor internally-- i.e., 16 bit data paths, >16 bit alu. But the 68000 is ALSO a 16 bit processor internally-- 16 bit >data path and alu. It just happens to have operations on longs built in... > >Not that I like the 8088, of course. > >[processor wars!] > -Miles [foot in mouth, you mean] The Motorola 68000 family of microprocessors ALL implement full 32-bit internal architecture. That means 32-bit registers, 32-bit data paths, and a 32-bit ALU. I recommend that you obtain FACTS next time instead of posting hearsay to the net. Try reading the Motorola MICROPROCESSOR COMPONENTS DATA MANUAL (or Hitachi's -- they're a second source) and then say "32 bits" until you're convinced. The 8088 is indeed a 16-bit machine (very strangely put together, it's true) which is then squished down onto an 8-bit data bus. Seems Intel wanted to make it `compatible' with the old 8-bit machines. An aside -- once some people (no names) with an Intel fixation tried to build a high-performance processor board. Two 8086's and an 8087 didn't do the job but a 68000 all by itself did. Tell you anything? ------------------------------- Dave Lewis Loral Instrumentation San Diego sdcc6 ---\ gould9 --\ ihnp4 ---->-->!sdcc3 ---->--->!loral!dml (uucp) sdcrdcf -/ sdcsvax -/ "Always put your clothes and weapons where you can find them quickly in the dark" -- Lazarus Long.
jlh@loral.UUCP (Desperatly seeking happiness) (12/24/85)
> > The 8088 is indeed a 16-bit machine (very strangely put together, it's true) > which is then squished down onto an 8-bit data bus. Seems Intel wanted to make > it `compatible' with the old 8-bit machines. > The 8088 is a 16 bit machine because IBM says it is. If you look in some of Intel's early data books (1978 or so) you'll see the 8088 referred to as an 8 bit processor. They didn't call it a 16 bitter until IBM needed to sell their pieces of junk, which would have been instantly recognized as junk had it been advertised correctly, i.e., having an 8 bit processor. Jim "When in trouble or in doubt run in circles, scream and shout"
jon@cit-vax.arpa (Jonathan P. Leech) (12/24/85)
> The Motorola 68000 family of microprocessors ALL implement full 32-bit > internal architecture. That means 32-bit registers, 32-bit data paths, and > a 32-bit ALU. I recommend that you obtain FACTS next time instead of posting > hearsay to the net. Try reading the Motorola MICROPROCESSOR COMPONENTS DATA > MANUAL (or Hitachi's -- they're a second source) and then say "32 bits" > until you're convinced. > > Dave Lewis Loral Instrumentation San Diego The 68000 has no 32 bit multiply or divide instructions. This does not seem like a `full 32-bit internal architecture' to me. This leads to an interesting question: should 68000 C compilers have sizeof(int) == 16 bits or 32 bits? The common choice is 32, I suspect so that poorly written VAX code would port without any work; but 16 bits seems more natural given the limitations of 32 bit arithmetic on the machine. Having a religious bias in favor of 68000s is fine - I have one, too - but don't let it blind you to FACTS. -- Jon Leech (jon@cit-vax.arpa) __@/
gwyn@brl-tgr.ARPA (Doug Gwyn <gwyn>) (12/24/85)
> This leads > to an interesting question: should 68000 C compilers have sizeof(int) > == 16 bits or 32 bits? Either is correct. Could we please stop arguing the relative merits of microprocessors on the C language newsgroup? Thanks.
farren@well.UUCP (Mike Farren) (12/26/85)
It seems to me that the arguments around 16-bit vs. 32-bit
architectures have gotten a little out-of-hand. There are two
DIFFERENT standards by which to judge processors. First, the width
of the physical data bus. By this standard, the 8088 is an 8-bit
processor, 8086 and 68000 16-bits, and the 68020 32-bits.
The second standard is the internal organization of the processor.
My own definition for this standard is the ability of the processor to
execute ALL of its data move, logical, and arithmetic instructions on
a given word size. ( Note. I emphasize that this is MY standard. ) By
this criterion, the Intel products up to and including the 80286 are ALL
16-bit processors, as are the 68000 and 68010. (They come REAL close to 32,
but fall down on the MUL instructions). The 68020 is a real 32-bitter in both
cases.
The point of all this is that there are several ways of looking at the
issue, any given one of which is correct, depending on what you're looking
for. I don't mind hearing discussions, even religious ones, but die-hard
dogmatism is never comfortable for those on the outside.
--
Mike Farren
uucp: {dual, hplabs}!well!farren
Fido: Sci-Fido, Fidonode 125/84, (415)655-0667
USnail: 390 Alcatraz Ave., Oakland, CA 94618hammond@petrus.UUCP (Rich A. Hammond) (12/27/85)
> Mike Farren writes: [ on 16 bit vs 32 bit] > The second standard is the internal organization of the processor. > My own definition for this standard is the ability of the processor to > execute ALL of its data move, logical, and arithmetic instructions on > a given word size. Of course, this puts RISC machines at an unfair advantage. For example the RISC I and II don't have a multiply or divide instruction, nor a complete single instruction 32 bit immediate mode (takes 2 instructions). Are they not 32 bit machines? If not, what are they? (13 bits, since that is the size of their immediate field for arithmetic? :-) ) If they are 32 bit machines, then shouldn't the 68000 and 68010, which can do all the add/shift/logical instructions the RISCs can for 32 bit operations, also be considered 32 bit machines? It's silly to classify a machine with a subset of the operations of another machine as being "wider" than the more complete instruction set machine. This gets even more complicated if you throw in floating point. :-) Rich Hammond [ihnp4|allegra|decvax|ucbvax] !bellcore!hammond
henry@utzoo.UUCP (Henry Spencer) (01/03/86)
> ... It's silly to classify > a machine with a subset of the operations of another machine as being > "wider" than the more complete instruction set machine. > > This gets even more complicated if you throw in floating point. :-) Even my poor little 11/44 is a 64-bit machine by some people's rules, since it has a 64-bit ALU in its floating-point processor. -- Henry Spencer @ U of Toronto Zoology {allegra,ihnp4,linus,decvax}!utzoo!henry
tim@ism780c.UUCP (Tim Smith) (01/04/86)
[ What is a 16 bit machine and what is an 8 bit machine? ] Don't forget the GI 1600! It has a 16 bit data bus, and a 10 bit instruction bus ( Yes, GI sold 10 bit wide roms ). What do we call this ( besides an abomination :-) ) ( by the way, 10 bits is called a decle, and 5 bits is a nickle ) ( GI also has one with a 13 bit instruction bus. That's progress. I don't know what one calls 13 bits ) -- Tim Smith sdcrdcf!ism780c!tim || ima!ism780!tim || ihnp4!cithep!tim
jsdy@hadron.UUCP (Joseph S. D. Yao) (01/05/86)
In article <217@ism780c.UUCP> tim@ism780c.UUCP (Tim Smith) writes: >( by the way, 10 bits is called a decle, and 5 bits is a nickle ) Shows you what inflation will do. When we were somewhat younger, 5 bits was $0.625. ;-) [And it isn't even anywhere near midnight!] -- Joe Yao hadron!jsdy@seismo.{CSS.GOV,ARPA,UUCP}
gwyn@BRL.ARPA (VLD/VMB) (01/06/86)
> 10 bits is called a decle, and 5 bits is a nickle > I don't know what one calls 13 bits Obviously, a trickle. CPU wordsize is not very interesting. I have software running on my 8-bit computer that beats anything some folks have on their 64-bit massively parallel computers (not for speed, but for utility and other human values).
bc10@rexago1.UUCP (Andrew C. Eisenberg) (01/06/86)
> ( by the way, 10 bits is called a decle, and 5 bits is a nickle ) > ( GI also has one with a 13 bit instruction bus. That's progress. > I don't know what one calls 13 bits ) > -- > Tim Smith sdcrdcf!ism780c!tim || ima!ism780!tim || ihnp4!cithep!tim Either a hypothyroid 8 bit machine, a midget 16 bit machine, or the result of someone's hangover. P.S. to Tim Smith for accidentally sending him a copy of this directly. -------------------------------------------------------------------------------- ________________ ________________ /___| | | | / []|| ROADWAY | | ROADWAY | |____||______________|-|______________| O-----O O O O Andrew C. Eisenberg -- rexago1!bc10@Case.CSNET - for you internet folks ...!decvax!cwruecmp!rexago1!bc10 - for UUCP "Beam me up Scotty, there is no intelligent life here!"
bright@dataioDataio.UUCP (Walter Bright) (01/07/86)
In article <217@ism780c.UUCP> tim@ism780c.UUCP (Tim Smith) writes: >( by the way, 10 bits is called a decle, and 5 bits is a nickle ) >( GI also has one with a 13 bit instruction bus. That's progress. > I don't know what one calls 13 bits ) Bad luck, obviously!
farren@well.UUCP (Mike Farren) (01/08/86)
In article <217@ism780c.UUCP>, tim@ism780c.UUCP (Tim Smith) writes: > ( by the way, 10 bits is called a decle, and 5 bits is a nickle ) > ( GI also has one with a 13 bit instruction bus. That's progress. > I don't know what one calls 13 bits ) Yet another opportunity for net innovation! If the abnormal fear of the number thirteen is "triskadecaphobia", then a 13-bit instruction must be a triscuit! (Yeah, I know this is a cheesy joke...) -- Mike Farren uucp: {dual, hplabs}!well!farren Fido: Sci-Fido, Fidonode 125/84, (415)655-0667 USnail: 390 Alcatraz Ave., Oakland, CA 94618