Path: utzoo!utgpu!attcan!uunet!husc6!mailrus!purdue!i.cc.purdue.edu!k.cc.purdue.edu!l.cc.purdue.edu!cik
From: cik@l.cc.purdue.edu (Herman Rubin)
Newsgroups: comp.arch
Subject: Re: A simple question on RISC
Summary: There are worse
Message-ID: <1013@l.cc.purdue.edu>
Date: 7 Nov 88 11:20:07 GMT
References: <76083@sun.uucp> <559@dms.UUCP> <1021@raspail.UUCP>
Organization: Purdue University Statistics Department
Lines: 26

In article <1021@raspail.UUCP>, bga@raspail.UUCP (Bruce Albrecht) writes:
> In article <559@dms.UUCP>, albaugh@dms.UUCP (Mike Albaugh) writes:
> > 	Oh, yeah, lack of Integer Mul & div was a bit of a pain, The
> > "Return Jump" a kluge (although a common one at the time), and One's
> > complement not exactly my cup of tea, so I don't maintain that the
> > 6600 was perfect, but considering the competition at the time...
> > 
> 
> The 6600 did have integer multiply (although the PP's did not).  I'm curious,
> though, are the Cybers the only machine around that simulate integer divide
> by doing a converting to floating point, floating point divide, normalize,
> and convert back to integer?

The integer multiply on the 6600 was an afterthought, and required a kludge. 
To avoid special cases, one either had to avoid the possibility of 48
significant bits in both factors, or go through conversion to floating
point.

As for divide, the CRAYs do not have any kind at all.  They do have a
reciprocal (floating point only) instruction which is accurate to a claimed
30 bits, and a correction factor instruction.  It is not necessarily true
that, say, 7n/n = 7.
-- 
Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907
Phone: (317)494-6054
hrubin@l.cc.purdue.edu (Internet, bitnet, UUCP)