Re: [802.3BA] 100GBASE-LR4 and ER4

Thread Links	Date Links
Thread Prev	Thread Next	Thread Index	Date Prev	Date Next	Date Index

Hi Alan,

Thank you for starting the discussion that we can hopefully continue during the informal Tuesday evening meeting next week.

Below are some of the trade-offs we went through in coming up with the numbers for the July presentations.

40GE 4x10G (cole_03_0708)

The Munich presentation (cole_03_0508) did not one have proposal but a range of numbers which bracketed the multiple proposals made by several companies during prior meetings. Our proposal was in cole_03_0308 and used the 10GE-LR link budget. TP2 and TP3 specifications have to take into account Mux and DeMux losses, and the 03_0308 proposal accounted for that mostly by increasing the transmitter output power.

The 03_0708 proposal reduced the link budget slightly, as pre-viewed in the 03_0308 presentation by taking advantage of today’s faster DFBs. It balanced the Mux and DeMux losses between the transmitter and receiver. The selected max TP2 power is about at the limit where un-cooled DFB eye safety limit control starts becoming an issue, and TIA saturation with very low loss DeMuxes starts becoming another issue. The resulting TP3 sensitivity has a lot of margin with respect to 10G receiver distribution. It is our view that the proposal is fully compatible with modern 10GE-LR un-cooled DFB transmitters and PIN receivers, and will re-use existing technology with ample manufacturing margin.

100GE 4x25G (cole_01_0708)

During the Q&A session after presentation of cole_01_0508 in Munich, several comments were made that the penalty allocation should be increased. We acknowledged that we were already looking at increasing the penalty allocations and would have an increase in a future proposal. This is along the lines of isono_01_0708, which also proposes a modest increase in penalty allocations. Cole_01_0708 increases the penalty allocations by 0.6dB from 01_0508. This 0.6dB is accommodated by a 0.2dB increase in transmitter output power and 0.4dB increase in receiver sensitivity. The resulting receiver sensitivity still has reasonable margin. Suggestions have been made to accommodate the link budget increase in other ways, and this can be subject of our future discussions.

I look forward to further discussion next week and over the summer to arrive at a consensus draft proposal for the September meeting.

Chris

From: Alan Tipper [mailto:alan.tipper@xxxxxxxxxxxxx]
Sent: Friday, July 11, 2008 10:44 AM
To: STDS-802-3-HSSG@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3BA] 100GBASE-LR4 and ER4

Pete,

‘refinements’ to 4x25G 10km baseline in COLE_01_0708 which increases Rx sensitivity by 0.4dB over that proposed in COLE_01_0508 which I agreed with.

Not 100G but 40G: The budget for 10km 4x10G COLE_03_0708 has -11.5dBm OMA Rx sensitivity per lane which is 1.1dB tougher than the Munich proposal COLE_03_0508 which I agreed with back in May. This seems to be taking us further away from 10G BASE LR technology re-use.

I know these numbers are subject to change at this stage but I’d like to understand the underlying rationale for the proposed changes in more depth than the time for Q&A in Denver would realistically allow.

Alan Tipper

Gennum Corp / Tacitis Ltd.

Alan.tipper@xxxxxxxxxxxxx

From: Pete Anslow [mailto:pja@xxxxxxxxxx]
Sent: 11 July 2008 18:08
To: Alan Tipper
Subject: RE: [802.3BA] 100GBASE-LR4 and ER4

Alan,

What were you concerned about?

Pete Anslow

Nortel Networks UK Limited, London Rd, Harlow, Essex CM17 9NA, UK

External +44 1279 402540 ESN 742 2540

Fax +44 1279 402543

From: Alan Tipper [mailto:alan.tipper@xxxxxxxxxxxxx]
Sent: 11 July 2008 18:07
To: Anslow, Pete (HAL02:Q840)
Subject: RE: [802.3BA] 100GBASE-LR4 and ER4

Hi Pete,

I’d certainly be interested in joining in an informal discussion over these specs – I was a bit concerned when going through the uploaded presentations.

See you in Denver.

Alan Tipper

Gennum Corp / Tacitis Ltd.

Alan.tipper@xxxxxxxxxxxxx

From: Pete Anslow [mailto:pja@xxxxxxxxxx]
Sent: 10 July 2008 23:27
To: STDS-802-3-HSSG@xxxxxxxxxxxxxxxxx
Subject: [802.3BA] 100GBASE-LR4 and ER4

Dear Task Force members,

During the process of discussing draft versions of contributions for the Denver meeting aimed at filling in the details of the 100GBASE-LR4 and ER4 optical parameter tables it has become clear that there are differing opinions over the detailed power budgets.

Since the Task Force has not yet adopted baseline presentations for all of its agreed objectives, it seems unlikely that there will be sufficient Task Force meeting time available next week to enter in to detailed discussion on optical parameter values for 100GBASE-LR4 and ER4. However, as the Editor of the clause containing these PMDs, I will be tasked with creating the first draft of it during the period between the Denver meeting and the September interim. I therefore wonder if there is any interest in holding an informal discussion to try to identify a set of parameter values that can be used as the basis of the first draft so that we can minimise the number of “TBDs” it contains.

If there is sufficient interest, I will confirm the time and location of the discussion during Tuesdays Task Force meeting.

See you all in Denver,

Pete Anslow

Nortel Networks UK Limited, London Rd, Harlow, Essex CM17 9NA, UK

External +44 1279 402540 ESN 742 2540

Fax +44 1279 402543