Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!samsung!uunet!csinc!rpeglar
From: rpeglar@csinc.UUCP (Rob Peglar)
Newsgroups: comp.arch
Subject: Re: F.P. vs. arbitrary-precision
Summary: History repeats itself.
Message-ID: <226@csinc.UUCP>
Date: 13 Sep 90 02:17:29 GMT
References: <3755@osc.COM> <4513@taux01.nsc.com> <119244@linus.mitre.org> <119612@linus.mitre.org>
Organization: Control Systems, Inc., St. Paul MN
Lines: 37

In article <119612@linus.mitre.org>, bs@linus.mitre.org (Robert D. Silverman) writes:
(bunch of stuff deleted)
>  
> The SPARC architecture did not use to have integer multiply and divide.
> They finally decided to add it. WHY? It is rumored that a certain
> gov't. agency complained so much about its lack that SUN had to put it
> in. Ask yourself what gov't. agency might require lots of multiprecise
> arithmetic. I do not know this for a fact, it is merely a rumor I heard.
> 

Isn't it funny how quite a bit of the uP world is rediscovering, reinventing,
rehashing, reliving, and redying all sorts of episodes that the MF and super
world went through, lo these many years ago?

"certain government agencies....forcing architectural changes..."

we've seen it all before.  no one should be surprised.

FP vs. arbitrary integer?  a *marketing* issue, not architectural.  If you
want to play the game, you have to know the rules.  In certain worlds, FP
rules.  In others, integer rules.  Pun intended.  Sure, there are many
clever ways to implement both.  Ya gotta sell it, though.  Note, no value
judgements here - enjoyable discussions.  It's just when people apply 
architectural prejudice on one or the other that the discussions become
absurd.  Re the coupla-months-ago thread about fork() vs. vfork().

Processors are fine.  Let's put our brains to work on memory and I/O,
the real hard issues.

Rob

uunet!csinc!rpeglar
-- 
Rob Peglar	Comtrol Corp.	2675 Patton Rd., St. Paul MN 55113
		A Control Systems Company	(800) 926-6876

...uunet!csinc!rpeglar