Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site brl-tgr.ARPA
Path: utzoo!linus!philabs!cmcl2!seismo!brl-tgr!tgr!GUBBINS@RADC-TOPS20.ARPA
From: GUBBINS@RADC-TOPS20.ARPA (Gern)
Newsgroups: net.micro
Subject: Re: NEC V20 ---> 8088
Message-ID: <1603@brl-tgr.ARPA>
Date: Thu, 19-Sep-85 09:51:39 EDT
Article-I.D.: brl-tgr.1603
Posted: Thu Sep 19 09:51:39 1985
Date-Received: Sat, 21-Sep-85 11:41:38 EDT
Sender: news@brl-tgr.ARPA
Lines: 43

Answers to your questions:

No, I do not have a NEC floppy disk controller.   The Z-100 (a completely
different, orginal design machine and far superior to the IBM/Clones)
uses a Western Digital 1797.

IBM liked the NEC because it was CHEAP.   Yes, I have several friends
that work at IBM that know these things as well as friends in the
Reliablity Division here (we are a USAF research and development lab)
that would have all sorts of nasty things to say about NEC chips, if
only they were allowed to say such things.

A just posted message has just confirmed (at least a rumor) that NEC
did indeed pirate the Intel 8088, right down to the early design mistake
in Intel's microcode (I did not think that the 8088 was a microcoded
CPU?).

Yes, I have a CMOS real-time clock calendar with auto-leap year in 
a device of my own design.    Working/interfacing with CMOS is a pain
DON'T TOUCH IT - IT DIES and the incompatible logic levels (pull-up,
pull-up...) and the slow speeds.   The very low power consumption
makes CMOS ideal for battery backup/battery powered applications such
as battery backup clock-calendar chips and the DG/One lap computer.
Sticking a CMOS CPU in a mostly TTL machine, especially in a borderline
spec designed machine such as the real IBM, may case line driving
problems.   Also the IBM design (again) is well known for its internal
heat problems and CMOS ICs age very rapidly as heat increases, much more
so than NMOS or TTL (in that order).   SO it may work now, but maybe
not a couple of years from now.  CMOS ICs that are ALMOST TTL level/
TTL drive level/TTL speed level do so at a large increase in power
consumption.    Granded, not as much as TTL, but the future is not
CMOS for high speed operations.  The inherent properties of CMOS
greatly limit its speed.

And I still hope Intel wins...

Cheers,
Gern

[The randoms in the message are my own opinions and usually that of
my friends and other sane persons too and may or may not be those of
the USAF, which won't tell]
-------