Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!caen!kuhub.cc.ukans.edu!1k1mgm
From: 1k1mgm@kuhub.cc.ukans.edu (Christopher Gunn)
Newsgroups: comp.unix.aix
Subject: Re: RS/6000 Multiuser Performance
Message-ID: <1991Mar29.112001.29354@kuhub.cc.ukans.edu>
Date: 29 Mar 91 11:20:01 CST
References: <27718@uflorida.cis.ufl.EDU>
Distribution: usa
Organization: KU Molecular Modeling
Lines: 30

In article <27718@uflorida.cis.ufl.EDU>, brs@cis.ufl.edu (Ray Seyfarth) writes:
> I am curious about experiences of users of RS/6000 model 320 systems
> with multiple users.  In particular I would like some feedback from
> users with about 30 X terminals attached to a 320.  What kind of
> performance is achieved and how much memory is needed?
> 
> Ray Seyfarth

I'm interested in this, too.  _Personal Workstation_ [nee _MIPS_] a
few months ago [January issue?  I thought I had it at my desk, but I
can't find it now....] had a long detailed series of articles on
the 320, one of which said they really sucked air doing serious
multi-user work.  I copied the articles and gave them to my IBM
salesman, who's looking into possible answers/rebuttals.  It was
unclear from the article whether the problem was inherent architecture,
cache size (which would be somewhat remedied by larger cache on xy0,
y >= 3), or AIX (which could at some point be fixed).

It's necessary to note that _Personal Workstation_ could better be
named _PC Hotrods_, and they may have some editorial bias in favor
of not giving their Everex 486-using readers bad cases of RS/6000-envy.
But the numbers they published were disturbing.

This may all have been discussed before I discovered that this
newsgroup was where the RS/6000 stuff got posted.  If so, I apologize
for bringing it up again....

Christopher Gunn	Molecular Graphics and Modeling Lab
SPAN--KUPHSX::GUNN	Department of Medicinal Chemistry, Malott Hall
913-864-4428 or -4495	University of Kansas, Lawrence, KS  66045