From caret@pyramid.com  Ukn Jul  4 00:28:38 1993
Received: from gossip.pyramid.com by SunSITE.unc.edu (4.1/tas-gen/1-30-93)
	id AA22818; Sun, 4 Jul 93 00:28:37 EDT
Received: from sword.eng.pyramid.com 
	by gossip.pyramid.com (5.61/OSx5.1a Pyramid-Internet-Gateway)
	id AA07457; Sat, 3 Jul 93 21:28:06 -0700
Received: by sword.eng.pyramid.com (5.61/Pyramid_Internal_Configuration)
	id AA17882; Sat, 3 Jul 93 21:27:54 -0700
Received: by sword.eng.pyramid.com (5.61/Pyramid_Internal_Configuration)
	id AA17875; Sat, 3 Jul 93 21:27:52 -0700
From: caret@pyramid.com (Neil Russell)
Message-Id: <9307040427.AA17875@sword.eng.pyramid.com>
Subject: Ethernet on the motherboard?
To: riscy@pyramid.com
Date: Sat, 3 Jul 93 21:27:52 PDT
X-Mailer: ELM [version 2.3 PL11]
Sender: riscy-request@pyramid.com
Reply-To: riscy@pyramid.com
Status: RO
X-Status: 

Opinion as to whether Ethernet is needed at all seems to be split
roughly 50-50.  For a good deal of us, this will be a super-fast
workstation in some small desk in the corner of our bedroom.
The only connection to the outside world would probably be a modem.
These people couldn't care less about Ethernet.  Others will connect
this to a small, or large network of workstations, and for them,
Ethernet is absolutely essential.  Others may use this as a diskless
workstation (read X-terminal), and then Ethernet is central.

The real question though is that if an ISA bus is provided, will
a motherboard solution provide so much extra gain in performance to
justify not using an ISA card?

Also, there are many problems getting a reliable Ethernet sub-system
(as some here have suggested).  Do we really want to get involved
with debugging Ethernet hardware?


I have two machines at home: a 386/33 and 486/33, connected by Ethernet.
I can get roughly 300 KB/sec transfer rate between them under the
right conditions.  I am using a WD8003E (Western Digital 8-bit card).
Does this mean that if I put a 16-bit card in I'll get 600KB/sec?
If so, this is getting close to the maximum data rate of Ethernet.
Is the ISA bus the bottleneck for Ethernet?

Assuming that a user's 'riscy' workstation is using motherboard video
and SCSI, then the only heavy user of the ISA bus left is the Ethernet
card.


There is enough bias in the above ramblings to anger anyone who doesn't
agree with me, which is exactly what I want.  Tell me why I'm wrong?
-- 
Neil Russell		(The wizard from OZ)
Pyramid Technology			Email:  caret@pyramid.com
3860 N. First Street			Voice:  (408) 428-7302
San Jose, CA 95134-1702			  FAX:  (408) 428-8845

 
From rei2!tsprad@uunet.UU.NET  Ukn Jul  4 17:26:10 1993
Received: from gossip.pyramid.com by SunSITE.unc.edu (4.1/tas-gen/1-30-93)
	id AA07416; Sun, 4 Jul 93 17:26:09 EDT
Received: from sword.eng.pyramid.com 
	by gossip.pyramid.com (5.61/OSx5.1a Pyramid-Internet-Gateway)
	id AA27118; Sun, 4 Jul 93 14:25:39 -0700
Received: by sword.eng.pyramid.com (5.61/Pyramid_Internal_Configuration)
	id AA29965; Sun, 4 Jul 93 14:25:26 -0700
Received: from goss.pyramid.com
	by sword.eng.pyramid.com (5.61/Pyramid_Internal_Configuration)
	id AA29956; Sun, 4 Jul 93 14:25:24 -0700
Received: from relay2.UU.NET 
	by gossip.pyramid.com (5.61/OSx5.1a Pyramid-Internet-Gateway)
	id AA27112; Sun, 4 Jul 93 14:25:31 -0700
Received: from spool.uu.net (via LOCALHOST) by relay2.UU.NET with SMTP 
	(5.61/UUNET-internet-primary) id AA21527; Sun, 4 Jul 93 17:24:04 -0400
Received: from rei2.UUCP by spool.uu.net with UUCP/RMAIL
	(queueing-rmail) id 172240.3873; Sun, 4 Jul 1993 17:22:40 EDT
Received: by rei.com (5.4/5.40/1.0)
	id AA16903; Sun, 4 Jul 1993 15:47:33 -0500
From: rei2!tsprad@uunet.UU.NET (6692)
Message-Id: <9307042047.AA16903@rei.com>
Subject: Ethernet on the motherboard?
To: riscy@pyramid.com
Date: Sun, 4 Jul 1993 15:47:32 -0500 (CDT)
In-Reply-To: <9307040722.AA15265@rei.com> from "UUCP Login" at Jul 4, 93 02:22:04 am
X-Mailer: ELM [version 2.4 PL13]
Content-Type: text
Content-Length: 1356      
Sender: riscy-request@pyramid.com
Reply-To: riscy@pyramid.com
Status: RO
X-Status: 

Neill said:
> 
> Opinion as to whether Ethernet is needed at all seems to be split
> roughly 50-50.  For a good deal of us, this will be a super-fast
> workstation in some small desk in the corner of our bedroom.
> The only connection to the outside world would probably be a modem.
> These people couldn't care less about Ethernet.  Others will connect
> this to a small, or large network of workstations, and for them,
> Ethernet is absolutely essential.  Others may use this as a diskless
> workstation (read X-terminal), and then Ethernet is central.
> 
> There is enough bias in the above ramblings to anger anyone who doesn't
> agree with me, which is exactly what I want.  Tell me why I'm wrong?

There is so little bias I can't tell if I agree or not.  I _think_
you're suggesting that Ethernet-on-board would be nice, but not as
easy as SCSI, and probably not worth the effort.  Since I can't
volunteer to design an Ethernet adapter (I am, after all, just a
software weenie) I have to agree:  Ethernet is nice but not
important to me.

-- 
Ted Spradley   Recognition International, Inc.  Opinions are mine, not theirs.
2701 E Grauwyler Rd. |Your productivity is not enhanced when you're staring
Irving TX 75061      |at that thing.  Your productivity is enhanced when
214-579-6692         |the computer is working and you're doing something else.

 
From tim@ubitrex.mb.ca  Ukn Jul  6 16:55:40 1993
Received: from gossip.pyramid.com by SunSITE.unc.edu (4.1/tas-gen/1-30-93)
	id AA18471; Tue, 6 Jul 93 16:55:39 EDT
Received: from sword.eng.pyramid.com 
	by gossip.pyramid.com (5.61/OSx5.1a Pyramid-Internet-Gateway)
	id AA02253; Tue, 6 Jul 93 13:55:05 -0700
Received: by sword.eng.pyramid.com (5.61/Pyramid_Internal_Configuration)
	id AA28537; Tue, 6 Jul 93 13:54:44 -0700
Received: from goss.pyramid.com
	by sword.eng.pyramid.com (5.61/Pyramid_Internal_Configuration)
	id AA28478; Tue, 6 Jul 93 13:54:40 -0700
Received: from ubiserver.ubitrex.mb.ca 
	by gossip.pyramid.com (5.61/OSx5.1a Pyramid-Internet-Gateway)
	id AA02237; Tue, 6 Jul 93 13:54:43 -0700
Received: from ska.ubitrex.mb.ca ([192.75.16.23]) by ubitrex.mb.ca (4.1/SMI-4.1)
	id AA10688; Tue, 6 Jul 93 15:53:34 CDT
Date: Tue, 6 Jul 93 15:53:34 CDT
From: tim@ubitrex.mb.ca (Tim Braun)
Message-Id: <9307062053.AA10688@ubitrex.mb.ca>
To: riscy@pyramid.com
Subject: Re: Ethernet on the motherboard?
Sender: riscy-request@pyramid.com
Reply-To: riscy@pyramid.com
Status: RO
X-Status: 

> The real question though is that if an ISA bus is provided, will
> a motherboard solution provide so much extra gain in performance to
> justify not using an ISA card?

I believe cost favours the m/board approach.

> Also, there are many problems getting a reliable Ethernet sub-system
> (as some here have suggested).  Do we really want to get involved
> with debugging Ethernet hardware?

> I have two machines at home: a 386/33 and 486/33, connected by Ethernet.

I think many of those initially involved will have an x86 based PC
box.  This is an argument in favor of ethernet on board, because 
it should allow using networking during the port.

> Assuming that a user's 'riscy' workstation is using motherboard video
> and SCSI, then the only heavy user of the ISA bus left is the Ethernet
> card.

Relying on ISA for ethernet opens configuration issues that we can avoid.

Another question is:

  10BaseT or 10Base2 or AUI?

I think the NatSemi DP8390 would give us good bang for our buck.  A dip 
version (48-pin DIP) is available.  Optional population (again) is a 
reasonable idea for the buyers, but is going to become intractible for 
the producer.

The DP83905/DP83902 from NatSemi has en/dec built in for ease of use with 
10BaseT.  10Base2 still requires a separate interface chip (DP8392).
 
________________________________________________________________
Tim Braun                          |
Ubitrex Corporation                | Voice: 204-942-2992 ext 228
1900-155 Carlton St                | FAX:   204-942-3001
Winnipeg, Manitoba, Canada R3C 3H8 | Email: tim@ubitrex.mb.ca

 
From drew@caesar.cs.Colorado.EDU  Ukn Jul  6 17:07:04 1993
Received: from gossip.pyramid.com by SunSITE.unc.edu (4.1/tas-gen/1-30-93)
	id AA19372; Tue, 6 Jul 93 17:06:59 EDT
Received: from sword.eng.pyramid.com 
	by gossip.pyramid.com (5.61/OSx5.1a Pyramid-Internet-Gateway)
	id AA03744; Tue, 6 Jul 93 14:06:12 -0700
Received: by sword.eng.pyramid.com (5.61/Pyramid_Internal_Configuration)
	id AA01764; Tue, 6 Jul 93 14:05:56 -0700
Received: from goss.pyramid.com
	by sword.eng.pyramid.com (5.61/Pyramid_Internal_Configuration)
	id AA01754; Tue, 6 Jul 93 14:05:53 -0700
Received: from caesar.cs.colorado.edu 
	by gossip.pyramid.com (5.61/OSx5.1a Pyramid-Internet-Gateway)
	id AA03736; Tue, 6 Jul 93 14:06:00 -0700
Received: from localhost by caesar.cs.Colorado.EDU with SMTP id AA02420
  (5.65c/IDA-1.4.4 for <riscy@pyramid.com>); Tue, 6 Jul 1993 15:05:45 -0600
Message-Id: <199307062105.AA02420@caesar.cs.Colorado.EDU>
To: riscy@pyramid.com
Subject: Re: Ethernet on the motherboard? 
In-Reply-To: Your message of "Tue, 06 Jul 1993 15:53:34 CDT."
             <9307062053.AA10688@ubitrex.mb.ca> 
Date: Tue, 06 Jul 1993 15:05:44 -0600
From: Drew Eckhardt <drew@caesar.cs.Colorado.EDU>
Sender: riscy-request@pyramid.com
Reply-To: riscy@pyramid.com
Status: RO
X-Status: 


    > The real question though is that if an ISA bus is provided, will
    > a motherboard solution provide so much extra gain in performance to
    > justify not using an ISA card?
    
    I believe cost favours the m/board approach.

Agreed.
    
    > Also, there are many problems getting a reliable Ethernet sub-system
    > (as some here have suggested).  Do we really want to get involved
    > with debugging Ethernet hardware?

Remember APPLICATION NOTES.  The sample implementation *will* 
work.
    
    Another question is:
    
      10BaseT or 10Base2 or AUI?

I'd vote for a Thinnet/AUI jumper selectable configuration.  Thinnet 
is cheap to wire, although trancievers are about $60 if you buy in small 
quantities, and it just takes another connector and a block of jumpers
to have AUI too.


    I think the NatSemi DP8390 would give us good bang for our buck.  

Agreed.  You think the dual-port memory approach is the way to go?

    A dip 
    version (48-pin DIP) is available.  Optional population (again) is a 
    reasonable idea for the buyers, but is going to become intractible for 
    the producer.

The Nat. Semi part 8390 is available in both PLCC and DIP parts, the former 
will use less realestate though.  

 
From tim@ubitrex.mb.ca  Ukn Jul  6 17:22:55 1993
Received: from gossip.pyramid.com by SunSITE.unc.edu (4.1/tas-gen/1-30-93)
	id AA20361; Tue, 6 Jul 93 17:22:54 EDT
Received: from sword.eng.pyramid.com 
	by gossip.pyramid.com (5.61/OSx5.1a Pyramid-Internet-Gateway)
	id AA05135; Tue, 6 Jul 93 14:21:31 -0700
Received: by sword.eng.pyramid.com (5.61/Pyramid_Internal_Configuration)
	id AA05432; Tue, 6 Jul 93 14:21:15 -0700
Received: from goss.pyramid.com
	by sword.eng.pyramid.com (5.61/Pyramid_Internal_Configuration)
	id AA05425; Tue, 6 Jul 93 14:21:13 -0700
Received: from ubiserver.ubitrex.mb.ca 
	by gossip.pyramid.com (5.61/OSx5.1a Pyramid-Internet-Gateway)
	id AA05131; Tue, 6 Jul 93 14:21:21 -0700
Received: from ska.ubitrex.mb.ca ([192.75.16.23]) by ubitrex.mb.ca (4.1/SMI-4.1)
	id AA10839; Tue, 6 Jul 93 16:20:13 CDT
Date: Tue, 6 Jul 93 16:20:13 CDT
From: tim@ubitrex.mb.ca (Tim Braun)
Message-Id: <9307062120.AA10839@ubitrex.mb.ca>
To: riscy@pyramid.com
Subject: Re: Ethernet on the motherboard?
Sender: riscy-request@pyramid.com
Reply-To: riscy@pyramid.com
Status: RO
X-Status: 

> From: Drew Eckhardt <drew@caesar.cs.Colorado.EDU>
>     
>     > Also, there are many problems getting a reliable Ethernet sub-system
  ...
> 
> Remember APPLICATION NOTES.  The sample implementation *will* 
> work.

It better.  It's the NE2000.
________________________________________________________________
Tim Braun                          |
Ubitrex Corporation                | Voice: 204-942-2992 ext 228
1900-155 Carlton St                | FAX:   204-942-3001
Winnipeg, Manitoba, Canada R3C 3H8 | Email: tim@ubitrex.mb.ca

 
