Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 12 Jan 2006 11:29:15 -0500
From:      Martin Cracauer <cracauer@cons.org>
To:        Danial Thom <danial_thom@yahoo.com>
Cc:        "Marc G. Fournier" <scrappy@hub.org>, Martin Cracauer <cracauer@cons.org>, freebsd-questions@freebsd.org
Subject:   Re: Dual Core vs HyperThreading vs Dual CPU
Message-ID:  <20060112112915.A16360@cons.org>
In-Reply-To: <20060112141925.91320.qmail@web33307.mail.mud.yahoo.com>; from danial_thom@yahoo.com on Thu, Jan 12, 2006 at 06:19:25AM -0800
References:  <20060111191224.A93090@cons.org> <20060112141925.91320.qmail@web33307.mail.mud.yahoo.com>

next in thread | previous in thread | raw e-mail | index | archive | help
> > I have benchmarks comparing dual-core 939
> > socket systems against dual
> > 940 socket systems here:
> >
> http://cracauer-forum.cons.org/forum/crabench.html
> 
> Just a question about your benches, any reason
> you just don't ship files to /dev/null? That was
> always the standard in unix to get the disk out
> of it.

I am not sure what specific test you are referring to, but in general
I avoid using /dev/null as a sink like the plague.  Speed on /dev/null
varies drastically between OSes, and sometimes between kernel versions
(e.g. SMP and non-SMP kernel).  In combination with other load it
becomes unpredictable.  Any benchmark doing dd of=/dev/null is bogus
IMHO. 

My cstream utility has a build-in disgard option which avoids this
problem.

Martin
-- 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Martin Cracauer <cracauer@cons.org>   http://www.cons.org/cracauer/
FreeBSD - where you want to go, today.      http://www.freebsd.org/



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060112112915.A16360>