Path: utzoo!attcan!uunet!snorkelwacker!apple!voder!nsc!amos
From: amos@nsc.nsc.com (Amos Shapir)
Newsgroups: comp.arch
Subject: Re: 64-bit code addresses
Message-ID: <13595@nsc.nsc.com>
Date: 14 Feb 90 00:55:35 GMT
References: <1926@l.cc.purdue.edu>
Organization: National Semiconductor, Santa Clara
Lines: 19
X-Hdate: 18 Shvat 5750

In article <1926@l.cc.purdue.edu> cik@l.cc.purdue.edu (Herman Rubin) writes:
>
>It is frequently the case that very long loops can be unrolled.

One has to be careful not to unroll too much, since that can lead to
unnecessary paging.

>Thus, code written by a human can easily translate
>into 2^20 or more instructions.

This is exactly my point:  We have barely scratched the surface of 32-bit
instruction space; 64-bit (or even 48) would probably never be needed
(certainly not within the same time frame of 64-bit data space, anyway).

-- 
	Amos Shapir
National Semiconductor, 2900 semiconductor Dr.
Santa Clara, CA 95052-8090  Mailstop E-280
amos@nsc.nsc.com or amos@taux01.nsc.com