Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 23 Apr 1998 18:06:25 -0700
From:      Don Wilde <dwilde1@ibm.net>
To:        Brian Behlendorf <brian@hyperreal.org>
Cc:        freebsd-advocacy@FreeBSD.ORG
Subject:   Re: [Fwd: Freeware]
Message-ID:  <353FE591.6A5A9F65@ibm.net>
References:  <353E34E3.308E0840@partsnow.com> <3.0.3.32.19980422184251.00af72e0@hyperreal.org> <3.0.3.32.19980422194339.00a32100@hyperreal.org> <3.0.3.32.19980423142234.008b9240@hyperreal.org>

next in thread | previous in thread | raw e-mail | index | archive | help
Brian Behlendorf wrote:

> >weekend. Would you be willing to advise such an effort insofar as
> >optimizing Apache setup?
> 
> Possibly, and I bet Dean Gaudet would be willing to join us.  His expertise
> is Linux but he'd be a help here as well.
> 
> >I haven't gotten any response from anybody official offering to take me
> >up on the challenge, so perhaps that is our best course: to proceed to
> >generate benchmarks in the normal SPEC way, by figuring out the ultimate
> >server we can manage to build. Let it be a learning experience if
> >nothing else, although I still have it in my head to approach some major
> >news organisation like 60 Minutes to sponsor the Challenge.
> 
> Heh.  Let's see what results we get first.
> 
> >Thanks for your offer, let's see what I can organize from our FreeBSD
> >side.
> 
> Great.  Are you a core FreeBSD developer?

>        Brian
> 
> 
> --=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--
> "Optimism is a strategy for making brian@apache.org
> a better future." - Noam Chomsky  brian@hyperreal.org
> 

No, I'm not. I'm a slightly-better-than-newbie user for 3 years, but
I've been programming real-time assembly language and Smalltalk for 15
years before that. I have gotten the official backing of FreeBSD-core in
this, and we're going to do it -- as a matter of fact, we're going to do
it _publicly_ -- and code-wise we will have all of their support. The
idea is that we will gather hardware and stage a media circus at
Berkeley, with all the historical connotations in full view.

As far as what it will take to win, it's pretty obvious that we're going
to have to get some corporate donations of hardware we can use for a
week or two to achieve these numbers. Did you look at the system Novell
used to make their numbers? It's posted on the SPECbench.org site. Only
one CPU (P-II 300), but FIVE 100Base-T cards and twenty PPro200's as
clients. The real secret is the cards they used are Intel SERVER cards,
and each has a 960 engine (I think) embedded on it with lots of FIFO.

There are other numbers for multiprocessor systems, but I think we want
to stick to one CPU. That'll be good enough, although it would be nice
to have a multiprocessor demo machine humming along doing something neat
like a Blender demo with Mesa and 3Dfx cards or something... 8-)))

--> Don



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-advocacy" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?353FE591.6A5A9F65>