Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!tut.cis.ohio-state.edu!ucbvax!ulysses!ulysses.att.com!cjc
From: cjc@ulysses.att.com (Chris Calabrese[mav])
Newsgroups: comp.sys.att
Subject: Re: Is the 3b2 dead?
Summary: floating point was never a strong point of the 3b
Keywords: support
Message-ID: <13119@ulysses.att.com>
Date: 14 Jun 90 13:24:08 GMT
References: <23609@adm.BRL.MIL> <286@alix.UUCP> <241@rossignol.Princeton.EDU> <917@gistdev.gist.com>
Sender: netnews@ulysses.att.com
Distribution: usa
Lines: 19

flint@gistdev.gist.com (Flint Pellett) writes:
> know) the software floating point emulation was performing sin() functions
> at the break-neck speed of 6 calls per second on an unloaded machine.  I

Face it, floating point on the 3b's always was terrible.  Why do you
need fast floating point on a phone switch?  By the time Summit
figured out that the thing wasn't being used as a phone switch any
more, it was too late to add decent floating point.  Same thing for
virtural memory hardare.  The original machines had fairly good VM for
swapping based systems, but it took a long time for them to get the
MMU right for paging.

All in all, they're not bad macines for general office automation type
use (the floating point's even pretty decent now).  They're just too
expensive for what they have to compete against.
Name:			Christopher J. Calabrese
Brain loaned to:	AT&T Bell Laboratories, Murray Hill, NJ
att!ulysses!cjc		cjc@ulysses.att.com
Obligatory Quote:	``Anyone who would tell you that would also try and sell you the Brooklyn Bridge.''