x86 architecture concepts

In article <lLITd.7192$OU1.4287@newssvr21.news.prodigy.com>,
	Kelly Hall <khall@acm.org> writes:
> Just out of curiosity, what are the dominant MCU architectures in the 
> embedded field?  Are there more elegant alternatives?  Why aren't they 
> dominant?

Hmm, I don't think there is any dominant MCU architecture in the embedded
market (at least not in the sense that the ia32 dominates the desktop
market). IMO, that the x86 is being used in embedded devices at all
is some kind of spillover-effect from the desktop market: a plethora
of tools and stuff available for cheap, etc. Interestingly, I would rate
the more successful architectures in the embedded market (e.g. ARM, PPC,
MIPS, 68k) to be all far more "elegant" than ia32. But that of course
depends on what one considers to be elegant. For me, it means achieving
a goal in a well thought-through way, without clumsy kludges or
work-arounds. That, applied to an MCU architecture apparently translates
to achieving a certain functionality with less transistors per square
millimeter which is the key reason why these architectures are successful
in the embedded market.

So, apparently, technical elegance does sometimes pay, *especially*
in the embedded market. This may be because the people making decisions
in the embedded market tend to be more technically knowledgable than
the average computer user  (just look at the people frequenting this
newsgroup!) and are not so easily caught by FUD campaigns.

Rob

-- 
Robert Kaiser                     email: rkaiser AT sysgo DOT com
SYSGO AG                          http://www.elinos.com
Klein-Winternheim / Germany       http://www.sysgo.com

Robert Kaiser wrote:
> So, apparently, technical elegance does sometimes pay, *especially*
> in the embedded market.

Bwahahahaha.  PIC.  8051.  Z8.  Forth.  RS-232.

Kelly

On 2005-02-26, Kelly Hall <khall@acm.org> wrote:
> Robert Kaiser wrote:
>> So, apparently, technical elegance does sometimes pay, *especially*
>> in the embedded market.
>
> Bwahahahaha.  PIC.

The PIC does seem like a bit of mess

> 8051.

When the 8051 was introduced, it was damned elegent.  

Everybody I knew loved it.  

Since then, people have taken it and tried to use it in some
very inappropriate (IMO) situations.  For example, the
architecture simply wasn't intended to deal with more than 64K
of code, with external RAM, or with a stack-intensive language
like C.  Sure there are kludges you can use to do those sorts
of things, but it's not pretty.

When used in the intended ways, the 8051 was (and is) rather
elegent.

Even the most elegent hammer in the world doesn't make a very
good screwdriver.

>  Z8. Forth.

Never used either.

> RS-232.

I've no complaints about RS-232.  I've seen it misused and
poorly implimented over the years, but I don't see anything
wrong with the standard.

-- 
Grant Edwards                   grante             Yow!  You should all JUMP
                                  at               UP AND DOWN for TWO HOURS
                               visi.com            while I decide on a NEW
                                                   CAREER!!

In article <vUYTd.945$C47.661@newssvr14.news.prodigy.com>,
	Kelly Hall <khall@acm.org> writes:
> Robert Kaiser wrote:
>> So, apparently, technical elegance does sometimes pay, *especially*
>> in the embedded market.
> PIC.  8051.  Z8.

I wouldn't know, never used any of these.

>  Forth.

What with Forth wrong is? Strange UPN you may find, but else?

>  RS-232.

And whats wrong with this?

Rob

-- 
Robert Kaiser                     email: rkaiser AT sysgo DOT com
SYSGO AG                          http://www.elinos.com
Klein-Winternheim / Germany       http://www.sysgo.com

Robert Kaiser wrote:
> In article <vUYTd.945$C47.661@newssvr14.news.prodigy.com>,
> 	Kelly Hall <khall@acm.org> writes:
> 
>>Robert Kaiser wrote:
>>
>>>So, apparently, technical elegance does sometimes pay, *especially*
>>>in the embedded market.
>>
>>PIC.  8051.  Z8.
> 
> 
> I wouldn't know, never used any of these.
> 
> 
>> Forth.
> 
> 
> What with Forth wrong is? Strange UPN you may find, but else?
> 
> 
>> RS-232.
> 
> 
> And whats wrong with this?
> 
It's hard to name anything that's right
with it.  Swinging several volts from
a positive rail to a negative rail to
transmit at a low symbol rate for a few
meters...

The 'standard' involved a DB-25 connector
with about a dozen signals, which everyone
implemented differently...

How many manhours has been wasted dicking
around with serial cables?

On Saturday, in article <8cSdnUye0uqibL3fRVn-hg@omsoft.com>
     jstewart@jkmicro.com "Jim Stewart" wrote:
>Robert Kaiser wrote:
>> In article <vUYTd.945$C47.661@newssvr14.news.prodigy.com>,
>>       Kelly Hall <khall@acm.org> writes:
>>>Robert Kaiser wrote:
>>>
>>>>So, apparently, technical elegance does sometimes pay, *especially*
>>>>in the embedded market.
>>>
...
>>> Forth.
>> 
>> 
>> What with Forth wrong is? Strange UPN you may find, but else?
>> 
>> 
>>> RS-232.
>> 
>> 
>> And whats wrong with this?
>> 
>It's hard to name anything that's right
>with it.  Swinging several volts from
>a positive rail to a negative rail to
>transmit at a low symbol rate for a few
>meters...

Having seen the standard being used to transmit down a couple
of kilometres, obviously I am wrong.

Unless of course you are referring to the MIS-interpreted
12-15m distance quote about noise.

>The 'standard' involved a DB-25 connector
>with about a dozen signals, which everyone
>implemented differently...

RS232 did NOT specify the conncetor until RS232-C onwards.
The DB-25 connector was a PTT standard (CCITT later ITU-T)
specification for modems, hence the DTE/DCE specifications.
Also the synchronous/asynchronous for 300baud to leased line
specifications.

RS232 did not specify things like DTR/DSR or CD/RI that was the
modem specifications hence NULL modem cables.

>How many manhours has been wasted dicking
>around with serial cables?

Mainly by those who did not understand the standard and grabbed
any old cable.

-- 
Paul Carpenter          | paul@pcserviceselectronics.co.uk
<http://www.pcserviceselectronics.co.uk/>    PC Services
<http://www.gnuh8.org.uk/>              GNU H8 & mailing list info
<http://www.badweb.org.uk/>             For those web sites you hate

Grant Edwards wrote:
> The PIC does seem like a bit of mess
> 
> 
>>8051.
> 
> 
> When the 8051 was introduced, it was damned elegent.  
> 
> Everybody I knew loved it.  
> 
> Since then, people have taken it and tried to use it in some
> very inappropriate (IMO) situations.  For example, the
> architecture simply wasn't intended to deal with more than 64K
> of code, with external RAM, or with a stack-intensive language
> like C.  Sure there are kludges you can use to do those sorts
> of things, but it's not pretty.
> 
It's been a long time since I last designed with and programmed a PIC 
(in Assembly) but I will comment on it all the same. AFAIK the PIC was 
derived from an IBM peripheral controller (IIRC PIC stands for 
peripheral intelligent controller) and it was meant to provide an 
additinal (yet minimal) level of intelligence, flexibility and 
compactness where usually plain logic would do fine. I think the problem 
with PICs is Microchip simply extended it far beyond the original 
architecture could stand and so they ended with a jurry rigged solution. 
Just as with 8x51. For that matter the x86 architecture suffers the same 
problem in order to support legacy applications.

Just my EU0.02.

On 2005-02-26, Elder Costa <elder.costa@terra.com.br> wrote:

> It's been a long time since I last designed with and
> programmed a PIC (in Assembly) but I will comment on it all
> the same. AFAIK the PIC was derived from an IBM peripheral
> controller (IIRC PIC stands for peripheral intelligent
> controller)

I think it was actually Peripheral Interface Controller.

> and it was meant to provide an additinal (yet minimal) level
> of intelligence, flexibility and compactness where usually
> plain logic would do fine.

I don't think it was IBM, but the rest is mostly right. General
Instrucments adapted the PIC from a Signetics design back in
the 70's to handle I/O related tasks for some other long since
extinct processor.  I think originally it came from Harvard or
somewhere.

> I think the problem with PICs is Microchip simply extended it
> far beyond the original architecture could stand and so they
> ended with a jurry rigged solution.  Just as with 8x51. For
> that matter the x86 architecture suffers the same problem in
> order to support legacy applications.

Well put.  The PIC, the 8051, and the IA32 have all suffered
from the universal hammer syndrome.

-- 
Grant Edwards                   grante             Yow!  Yow! Those people
                                  at               look exactly like Donnie
                               visi.com            and Marie Osmond!!

On 26 Feb 2005 20:54:26 +0100, rob@rupert.sysgo.com (Robert Kaiser)
wrote:

>
>>  RS-232.
>
>And whats wrong with this?

While the RS-232 "standard" may be adequate for the initial purpose of
connecting a serial signal from a computer (DTE) to a modem (DCE) in
an adjacent rack, the worst problems with RS-232 for any general
purpose interface is the far too high impedance levels compared to any
realistic cable characteristic impedance and the problem with ground
potential differences, since no galvanic isolation is required.

Paul
 

In article <opsmnukld2y1ubid@news>,
Vadim Borshchev  <vadim.borshchev@127.0.0.1> wrote:
>On Wed, 23 Feb 2005 17:27:55 +0100, Mouarf <mouarf@chezmoi.fr> wrote:
>
>> I'm not sure, see the PXA processor from Intel that use ARM
>> architecture........
>
>Absolutely.  PXA is based on XScale (micro)architecture, which is in turn 
>"ARM Architecture 5TE", by ARM nomenclature.
>
>Your original question was about x86, which is completely different can of 
>worms.  It all has started the day Intel decided to ..ermm.. "extend" 
>their i8080 processor and make it 16-bit.  The new i8086 processor had 
>inherited the odd registers functionality and names.  Subsequent family 
>members (286, 386, 486, 586 aka Pentium, 686 aka PentiumII) just added to 
>the mess.  Due to the numbering convention this whole line is referred to 
>as "x86".

I have the original introduction by Morse the designer of the
8086. He is somewhat critical of his own design and Intels assembler,
too. I think the 8086 is a decent processor, and a feat of upwards
compatibility. But it really should have been the end of it. 
Extending the lifetime of the processor from 16 bits to 20 bits
addressing, counting for 4 years in computer history, and there it
should have stopped.

 The real mess didn't start until the addition of protected mode.
Even so, some Unix clones (notably Coherent) show that there was 
a way out.

But Microsoft killed the attempts of IBM to get out of the BIOS
morass (microchannel) and forced Intel to introduce the "virtual
real mode". 

Even so Linux proves that if you ignore sufficient of its features
the Intel 386 is a decent processor. A great deal of the mess
in a Linux system has to do with the battle for the control of the
boot code waged by Microsoft (BIOS).

>As Intel produces Itanium (or do they still?), the "x86" get another name, 
>"IA32".  Itanium, for it's 64-bit nature, is named "IA64".
>
>And now ask somebody about i8035 or i8051 :)

8051 another dead-ugly processor ...
But I prefer them over the PIC's.

>
>
>HTH,
>   Vadim



-- 
-- 
Albert van der Horst,Oranjestr 8,3511 RA UTRECHT,THE NETHERLANDS
        One man-hour to invent,
                One man-week to implement,
                        One lawyer-year to patent.