Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!munnari.OZ.AU!news.ecn.uoknor.edu!feed1.news.erols.com!howland.erols.net!newsxfer3.itd.umich.edu!news1.best.com!nntp1.ba.best.com!not-for-mail
From: dillon@flea.best.net (Matt Dillon)
Newsgroups: comp.unix.sco.misc,comp.unix.bsd.freebsd.misc,comp.unix.bsd.bsdi.misc,comp.sys.sgi.misc
Subject: Re: no such thing as a "general user community"
Date: 25 Mar 1997 16:59:42 -0800
Organization: BEST Internet Communications, Inc.
Lines: 33
Message-ID: <5h9sdu$3ag@flea.best.net>
References: <331BB7DD.28EC@net5.net> <5glbkq$io6$3@kayrad.ziplink.net> <5gn615$9cu@flea.best.net> <5h92no$gua@innocence.interface-business.de>
NNTP-Posting-Host: flea.best.net
Xref: euryale.cc.adfa.oz.au comp.unix.sco.misc:37310 comp.unix.bsd.freebsd.misc:37717 comp.unix.bsd.bsdi.misc:6457 comp.sys.sgi.misc:29397
:In article <5h92no$gua@innocence.interface-business.de>,
:J Wunsch <joerg_wunsch@interface-business.de> wrote:
:>dillon@flea.best.net (Matt Dillon) wrote:
:>
:>> A disk configuration is a disk configuration no matter what
:>> platform you are running on. Memory capacity is memory
:>> capacity.. the memory requirements for a particular program are
:...
:>instructions in order to load a single address, and that all addresses
:>are 64-bit there, that you have more wasted space by padding etc. But
:>it simply shows that the memory requirements of the same task on an
:>Alpha CPU are very likely to be higher than they were on a 32-bit CISC
:>CPU.
:>
:>--
:>J"org Wunsch Unix support engineer
:>joerg_wunsch@interface-business.de http://www.interface-business.de/~j
Mmm.. respectfully, I would disagree. This might be true for very large
binaries, but for most programs the DATA+RSS+ALLOCATION will completely
override the size of the TEXT area (portion of the binary that holds
the instructions). The only time I see big differences in a ps is with
shared libraries, which vary greatly across architectures, but being
shared they do not have much impact on the true memory use.
I think I *have* noticed differences in L1 and L2 cache effectiveness.
RISC boxes tend to need more L1 and L2 cache due to the larger
instructions. It's interesting to note that Intel can get away with
a measily 256K or 512K L2 cache on its PPROs. For all of its complexity,
the x86/386/486/pentium instruction set is pretty compact.
-Matt