From owner-freebsd-hackers  Thu May 18 12:08:57 1995
Return-Path: hackers-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id MAA15971
          for hackers-outgoing; Thu, 18 May 1995 12:08:57 -0700
Received: from gndrsh.aac.dev.com (gndrsh.aac.dev.com [198.145.92.241])
          by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id MAA15965
          for <hackers@FreeBSD.org>; Thu, 18 May 1995 12:08:53 -0700
Received: (from rgrimes@localhost) by gndrsh.aac.dev.com (8.6.11/8.6.9) id MAA14860; Thu, 18 May 1995 12:07:18 -0700
From: "Rodney W. Grimes" <rgrimes@gndrsh.aac.dev.com>
Message-Id: <199505181907.MAA14860@gndrsh.aac.dev.com>
Subject: Re: ram-speed (was Re: Adaptec 2940?)
To: adhir@iagi.net (Alok K. Dhir)
Date: Thu, 18 May 1995 12:07:17 -0700 (PDT)
Cc: aledm@pavilion.co.uk, hackers@FreeBSD.org
In-Reply-To: <Pine.BSI.3.91.950518111626.21857B-100000@bigdipper.iagi.net> from "Alok K. Dhir" at May 18, 95 11:19:26 am
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1126      
Sender: hackers-owner@FreeBSD.org
Precedence: bulk

> 
> 
> Some more data points:
> 
> Bigdipper 	(ASUS SP3G, 486dx2/66, 40mb RAM, 512k cache, 1.1.5.1)
> 
> 	49005fb0   0.309 uS/op 3.23e+06 op/S 12.327 Mb/S
> 	8938c0df   0.159 uS/op 6.31e+06 op/S 24.066 Mb/S
> 
> 
> Littledipper 	(ASUS SP3G, 486dx2/66, 32mb RAM, 512k cache, -current)
> 
> 	49005fb0   0.312 uS/op 3.21e+06 op/S 12.233 Mb/S
> 	8938c0df   0.158 uS/op 6.32e+06 op/S 24.112 Mb/S

I've been telling these folks that the ASUS PCI/I-486SP3G was one
heck of a fast motherboard, and here are the numbers from an independent
source to prove it for me!!

Those memory numbers are as fast as the Neptune P90 ASUS board, kinda
pisses me off that I can't get 48MB/sec out of the Neptune when I
can get 24MB/sec out of the Saturn II!

> (BTW - both machines were pretty heavily loaded at the time...  Not sure 
> if it makes a difference)

It does not as long as there is the free 8MB or so of memory needed
by this process (ie, page faulting would rune the results).


-- 
Rod Grimes                                      rgrimes@gndrsh.aac.dev.com
Accurate Automation Company                   Custom computers for FreeBSD