From owner-freebsd-hardware  Fri Jun 28 12:57:38 1996
Return-Path: owner-hardware
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id MAA25194
          for hardware-outgoing; Fri, 28 Jun 1996 12:57:38 -0700 (PDT)
Received: from jparnas.cybercom.net (jparnas.cybercom.net [206.28.135.58])
          by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id MAA25186
          for <freebsd-hardware@freebsd.org>; Fri, 28 Jun 1996 12:57:35 -0700 (PDT)
Received: from localhost.cybercom.net (localhost.cybercom.net [127.0.0.1]) by jparnas.cybercom.net (8.6.10/8.6.10) with SMTP id PAA04592; Fri, 28 Jun 1996 15:56:21 -0400
Message-Id: <199606281956.PAA04592@jparnas.cybercom.net>
X-Authentication-Warning: jparnas.cybercom.net: Host localhost.cybercom.net didn't use HELO protocol
To: Chuck Robey <chuckr@glue.umd.edu>
cc: Kevin Swanson <Kevin_Swanson@blacksmith.com>, freebsd-hardware@freebsd.org
X-External-Networks: yes
Subject: Re: muliport boards - building a PPP dialup server 
In-reply-to: Your message of Fri, 28 Jun 1996 15:45:25 EDT.
             <Pine.OSF.3.91.960628154114.1057E-100000@thurston.eng.umd.edu> 
Date: Fri, 28 Jun 1996 15:56:19 -0400
From: "Jacob M. Parnas" <jparnas@jparnas.cybercom.net>
Sender: owner-hardware@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk


In message <Pine.OSF.3.91.960628154114.1057E-100000@thurston.eng.umd.edu>you write:
>On Fri, 28 Jun 1996, Jacob M. Parnas wrote:
>
>> 
>> In message <Pine.OSF.3.91.960628143531.1057C-100000@thurston.eng.umd.edu>you write:
>> >On Fri, 28 Jun 1996, Kevin Swanson wrote:
>> >
>> >> I was told by Digi that their "dumb" 8-port boards (PC/8), could only go up  
>> >> to 19.2 KBps if you used 16550's. This doesn't sound right based on some of  
>> >> the responses I've been getting.
>> 
>> These days, a multiboard should at least do 115200baud in my opinion.  At
>> least it can be competitive with modems.  8x115200baud is still only 1/8 
>> MB/sec and 16x115200 is only 1/4 MByte/sec.  Not exactly fast.
>
>You notice he referred to the "dumb" 8 port digiboards?  They were much 
>cheaper than the intelligent ones, and serve many applications perfectly 
>well, but since they have no onboard intelligence or buffering (beyond 
>that allowed by the 16550 fifos) they won't make it to 115200, at least 
>not on all 8 ports at the same time.

I think "dumb" ones should be able to go at least 115200 per port (for even
$120-$130 new modems.  Smart ones should support at least 512kbaud/port.)
Some like ones based on the TI 17550 chip (I think) go up to > 900 Kbaud/sec.

Jacob