f



Processor Generations

Can anyone tell me what processor Generation is Intel at (seven, eight)?

Thanks
0
itsski2
6/30/2003 6:35:44 PM
comp.arch 7611 articles. 0 followers. carchreader (32) is leader. Post Follow

5 Replies
1132 Views

Similar Articles

[PageSpeed] 59

mailer-daemon@bof.de (Patrick Schaaf) wrote in message news:<3f008464$0$16583$9b622d9e@news.freenet.de>...
> itsski2@yahoo.com (Ski) writes:
> 
> >Can anyone tell me what processor Generation is Intel at (seven, eight)?
> 
> Do you know what a complex number is?
> 
> SCNR
>   Patrick

Maybe that should be an irrational number?

- Jeff Dutky
0
dutky
6/30/2003 9:14:55 PM
Patrick Schaaf wrote:

> itsski2@yahoo.com (Ski) writes:
> 
>>Can anyone tell me what processor Generation is Intel at (seven, eight)?
> 
> Do you know what a complex number is?

Intel has a disruptive numbering. They started with the 4004, followed by 
the 8008 and 8080. The 8085 is small derivative of the 8080, and I'd 
considder it as the same (third generation) processor. The 8086 was the 
follow-on, thus the fourth generation (or the sixth, if the .6 makes any 
sense ;-). However, the 8086 also resets the counter, so we have 8086, 
80286, 80386, 486 (yes, without 80!), i.e. 486 is fourth generation 8086 or 
seventh generation of CPUs at Intel. Then came another disruptive counting, 
the Pentium. We have Pentium, Pentium MMX, PentiumPro (P6), Pentium II, 
Pentium III and Pentium 4 (not IV!). Pentium and Pentium MMX are the same 
architecture, PentiumPro, II and III, too. Pentium 4 is the third 
architecture after Pentium or the tenth CPU generation at Intel. And then, 
there's the Penium M. It's inspired by the P6 architecture, but has 
sufficient differences to be called a new architecture. Is it eleventh? No, 
from a roadmap point of view, the Pentium M is not meant as competition to 
the P4 (though it needs a 2.8GHz P4 to beat a 1.6GHz Pentium M), but as 
low-power offering for a different market (notebooks and blades). There are 
now two quite different CPUs of the same generation available as "Pentium".

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/
0
Bernd
6/30/2003 10:25:30 PM
On Tue, 1 Jul 2003, Bernd Paysan wrote:

> sense ;-). However, the 8086 also resets the counter, so we have 8086,
> 80286, 80386, 486 (yes, without 80!), i.e. 486 is fourth generation 8086 or

80188/80186.

-Peter

"We did not start this war," Rumsfeld said. "So understand, responsibility for every single
casualty in this war, whether they're innocent Afhgans or innocent Americans, rests at the
feet of al-Qaeda and the Taliban."
0
Peter
6/30/2003 11:50:37 PM
Paul Wallich wrote:
> You're missing the 80186 there, which was a pile-more-stuff-on-the-die 
> version of the 8086, but still could be considered a generation insofar 
> as moving stuff onto the CPU chip was considered an architectural 
> decision...

I'd say the 186 was a bugfix for the original 8086:

A few important opcodes made it onto the die, no longer any problem with 
having an interrupt in the middle of reloading the SS:SP stack pair.

Terje

-- 
- <Terje.Mathisen@hda.hydro.com>
"almost all programming can be viewed as an exercise in caching"

0
Terje
7/1/2003 6:47:13 PM
Terje Mathisen <terje.mathisen@hda.hydro.com> wrote in message news:<bdskvc$8qt$1@osl016lin.hda.hydro.com>...
> I'd say the 186 was a bugfix for the original 8086:
> 
> A few important opcodes made it onto the die, no longer any problem with 
> having an interrupt in the middle of reloading the SS:SP stack pair.


More like a new (or significantly enhanced) core, plus a lot of
integrated peripherals.  Most of the non-trivial instructions were at
least a few clocks faster, some of them substantially so (multiply,
divide, string handling, rotates, etc.).  And a few things were slower
(in a couple of unusual cases, significantly (for example, AAA on the
186 took twice as many clocks, while AAM ran four times as fast).

Unfortunately the integrated stuff wasn't IBM PC compatible so the 186
was rarely used in PCs.  Although it did score about a million
embedded design wins.
0
robertwessel2
7/2/2003 5:57:58 AM
Reply: