Path: utzoo!attcan!utgpu!watmath!att!tut.cis.ohio-state.edu!purdue!mailrus!cornell!uw-beaver!fluke!jfadams
From: jfadams@tc.fluke.COM (Jim Adams)
Newsgroups: comp.graphics
Subject: Re: CACHE on 386 and GRAPHICS PERFORMANCE
Message-ID: <11268@fluke.COM>
Date: 21 Sep 89 14:50:49 GMT
References: <12719@s.ms.uky.edu>
Sender: news@tc.fluke.COM
Organization: John Fluke Mfg. Co., Inc., Everett, WA
Lines: 17

I use a Compaq 386/20e at work and built a 386 clone at home. The Compaq
runs at ~20MHz with a cache and the homebrew at 25MHz without one. They
both have 22ms Hard disks and 387 coprocessors.  I cannot distinguish any
difference in performance between the two machines even when running
graphics-rich applications such as CAD programs and DTP programs.  I would
sink my $ in a FAST disk drive.

In addition, some clone boards provide memory interleaving when all RAM
banks are filled, providing (for me, anyway) an increase in performance from
an equilavence of 25MHz to 31MHz.

-- Jim Adams
-- 
James F. Adams       John Fluke Mfg. Co., Inc.        Everett, Washington USA
WORLD:jfadams@tc.fluke.COM
UUCP:{ihnp4!uw-beaver,ucbvax!lbl-csam,allegra,decvax!microsoft}!fluke!jfadams
ARPA:fluke!jfadams@uw-beaver.ARPA  GEnie:J.F.ADAMS  CIS:74036,2517