80286, flaming and Multics
Phil Ronzone
phil at Apple.COM
Sat Apr 8 03:39:36 AEST 1989
In article <1188 at frog.UUCP> john at frog.UUCP (John Woods) writes:
>In article <18878 at adm.BRL.MIL>, dana at bilbo.locus (Dana Myers) writes:
>> The 80286 was designed by people who liked Multics and figured they would
>> make the perfect microprocessor to run Multics on.
>
>These people were doing some really heavy-duty drugs at the time, right?
>Or were they just inhaling raw gasoline?
I worked for Bill Pohlman, known as the father of the 286, after he left Intel.
We discussed how the whole xxx86 architecture came about.
To be overly brief, the xxx86 architecture started as a stopgap. Intel believed
the 432 was the wave of the future, yet the Intel salesman were VERY worried
that that the Zilog Z8000 would kill Intel before Intel could get a real
432 chip set out the door (remember what the Z80 did to the 8080?).
Anyway, the whole xxx86 was subject to severe constraints to keep it from
being "too good" and competing with the 432 (no floating point plans etc.).
To keep the xxx86 alive, it was touted as an upgrade path for those old 8080
customers, who were eventually switch to to the 432.
Bill mentioned that the 8088 was done almost as an afterthought, and was
almost cancelled several times.
Bill also has an unpublished paper titled "Why the father of the 80286 thinks
the Motorola 68000 is better", but it remains unpublished because Bill didn't
want to burn any bridges at Intel. Last I heard, he was back at Intel in Paris.
+------------------------+-----------------------+----------------------------+
| Philip K. Ronzone | A/UX System Architect | APPLELINK: RONZONE1 |
| Apple Computer MS 27AJ +-----------------------+----------------------------+
| 10500 N. DeAnza Blvd. | Computer viruses don't cause security problems, |
| Cupertino CA 95014 | computer programmers do ... |
+------------------------+----------------------------------------------------+
|{amdahl,decwrl,sun,voder,nsc,mtxinu,dual,unisoft}!apple!phil |
+-----------------------------------------------------------------------------+
More information about the Comp.unix.wizards
mailing list