From owner-freebsd-current  Tue May  7 17:08:15 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id RAA27723
          for current-outgoing; Tue, 7 May 1996 17:08:15 -0700 (PDT)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id RAA27714
          for <current@freebsd.org>; Tue, 7 May 1996 17:08:11 -0700 (PDT)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.12/8.6.9) id KAA12703; Wed, 8 May 1996 10:02:20 +1000
Date: Wed, 8 May 1996 10:02:20 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199605080002.KAA12703@godzilla.zeta.org.au>
To: rgrimes@GndRsh.aac.dev.com, wscott@ichips.intel.com
Subject: Re: more on fast bcopy
Cc: asami@cs.berkeley.edu, bde@zeta.org.au, culler@cs.berkeley.edu,
        current@freebsd.org, ken@area238.residence.gatech.edu, marc@bowtie.nl,
        nisha@cs.berkeley.edu, pattrsn@cs.berkeley.edu, wollman@lcs.mit.edu
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

>>     size    bandwidth
>>...
>>     4096  69.565217 MB/s
>>     8192  71.111111 MB/s

>Thats pretty cruddy for 4 way interleaved memory,
>1/120nS * 8 bytes/leave * 4 leaves == 266MB/sec.

Erm, divide by 2 for read+write.

>And I used the worst case there by using the 120nS full cycle time,
>if these are sequentail accesses that hit the same DRAM page this
>number should be up closer to 500MB/sec.

Mabye this accounts for some of the advantages of the FP method - it's
easy to read 8*8 bytes off a page at a time, while using integer
registers it's impossible to read more than 5*4 bytes at a time (even
using %ebp and %esp as scratch registers :-).

Bruce