From owner-freebsd-questions  Wed Mar 26 14:36:08 1997
Return-Path: <owner-questions>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.5/8.8.5) id OAA05327
          for questions-outgoing; Wed, 26 Mar 1997 14:36:08 -0800 (PST)
Received: from luke.cpl.net ([206.85.245.131])
          by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA05316
          for <questions@freebsd.org>; Wed, 26 Mar 1997 14:36:05 -0800 (PST)
Received: from localhost (shawn@localhost) by luke.cpl.net (8.8.5/8.7.3) with SMTP id OAA03582; Wed, 26 Mar 1997 14:35:39 GMT
Date: Wed, 26 Mar 1997 14:35:39 +0000 (GMT)
From: Shawn Ramsey <shawn@luke.cpl.net>
To: Ben Black <black@zen.cypher.net>
cc: "Paul T. Root" <proot@horton.iaces.com>, questions@freebsd.org
Subject: Re: SCSI-II vs Ultra-SCSI
In-Reply-To: <Pine.LNX.3.91.970326172507.27209F-100000@zen.cypher.net>
Message-ID: <Pine.BSF.3.95.970326143416.3561B-100000@luke.cpl.net>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-questions@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

> > performance than the 1542B. I have a Gigabyte MB with Triton II with
> > the 2 EIDE controllers.
> > 
> 
> this is the usual argument against scsi, but it is simply wrong.  they 
> key to scsi is that the controller and devices are intelligent.  they 
> handle a lot of the I/O processing on their own *without* CPU 
> intervention.  EIDE requires CPU assistance for just about everything.  
> if you have a lot of I/O on your machine, performance will suffer because 
> the CPU will be handling so many of the EIDE interrupts.

Isnt that only true of Busmaster SCSI controllers? There are PIO SCSI
controllers, although im not sure if that one is. (Note I am not saying he
should use IDE, by any means. Just get a better controller. :) )