Re: [802.3_100GNGOPTX] MMF Ad Hoc - 29th Nov meeting notes, and next meeting date

[Daniel Dove <ddove@xxxxxxx>]

Hi Jonathan,

Thanks for driving this.

I am still a big fan of breaking out the various cost elements into a stacked bar-chart based on the reach capabilities of the various combinations, then laying that against cumulative distribution data for server/switch and switch/switch links. It just seems like that would give a useful perspective even if the data used to generate it was somewhat subjective.

The paper referenced below, and a few others I found cited in that paper provide some interesting reading. It seems that higher density leads to lower TCO and that there are finite limitations to the number of nodes you can cluster within a reasonable space, thus leading me to believe that there will be a large volume of short-reach links. (<30m…maybe even <25m).

If so, then the natural follow-up question is whether there is a significant cost delta between a 30m PMD (server/switch or mid-rack to cluster-switch) and a PMD intended to satisfy the longer cluster-cluster distribution links that are likely to run up to 100m. 

Dan

From: Jonathan King <jonathan.king@xxxxxxxxxxx>
Date: Wed, 30 Nov 2011 15:38:30 -0800
To: 100G Group <STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx>
Cc: Alan Flatman <a_flatman@xxxxxxxxxxxxx>, "Amezcua, A. (Adrian)" <adrian.amezcua@xxxxxxxxx>, Peter Anslow <panslow@xxxxxxxxx>, Anthony Torza <anthony.torza@xxxxxxxxxx>, "Bernstein, Gary" <GBernstein@xxxxxxxxxxx>, Brad Booth <Brad_Booth@xxxxxxxx>, Daniel Dove <ddove@xxxxxxx>, Ephrem Wu <ephremw@xxxxxxxxxx>, Gary Nicholl <gnicholl@xxxxxxxxx>, Harry Fu <harryf@xxxxxxxxxx>, Jack Jewell <jack@xxxxxxxxxxxxxx>, Jeffery Maki <jmaki@xxxxxxxxxxx>, John Petrilla <john.petrilla@xxxxxxxxxxxxx>, Keith Nellis <knellis@xxxxxxxxx>, "Kolesar, Paul" <PKOLESAR@xxxxxxxxxxxxx>, Lian Zhao <lzhao@xxxxxxxxx>, Martin Gilpatric <marting@xxxxxxxxxx>, <mike.dudek@xxxxxxxxxx>, Mark Nowell <mnowell@xxxxxxxxx>, Oren Sela <orens@xxxxxxxxxxxx>, Piers Dawe <piers.dawe@xxxxxxxxxxxxx>, Robert Coenen <rcoenen@xxxxxxxxxxxxxxxxxxx>, Ryan Latchman <ryan.latchman@xxxxxxxxxxxxx>, Scott Kipp <skipp@xxxxxxxxxxx>, Shmuel Levy <shmuel@xxxxxxxxxxxx>, Sudeep Bhoja <sbhoja@xxxxxxxxxxxx>, Vipul Bhatt <Vipul_Bhatt@xxxxxxxx>
Subject: RE: [802.3_100GNGOPTX] MMF Ad Hoc - 29th Nov meeting notes, and next meeting date

Dear all

1)      Here are draft  unapproved meeting notes for the MMF ad hoc meeting of 29^th November.

Please send me suggestions for changes/corrections or comments !

 

2)      By popular demand, we will NOT meet on 6^th December, so the next meeting will be on Tuesday 13^th December, 10am – 11 am PCT.

 

3)      I’d like to encourage you all to continue discussions on the reflector.  There’s been some traffic on the paper recommended to us by Andy Moorwood:

“The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines

Luiz André Barroso and Urs Hölzle”

www.morganclaypool.com

ISBN: 9781598295566 paperback

ISBN: 9781598295573 ebook

The paper may be viewed, without charge, at the Morgan & Claypool site, but please be aware of the restrictive notices on page (iv)    http://www.morganclaypool.com/doi/pdf/10.2200/S00193ED1V01Y200905CAC006

Anyone care to put together a summary of the sections that are relevant to the MMF ad hoc ? pretty please?

4)      Paul Kolesar has corrected a couple if typos in the slides he presented on 29^th Nov (date code and his affiliations logo), a corrected slide set will be posted to the MMF ad hoc materials page (thanks Kapil)

 

Thanks for your involvement !

Best wishes

Jonathan