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

Re: [802.3_B400G] adopted objectives



An off-line thread reminded me that I did not cite presentations by Adee Ran which initiated the subject which I commented on in my last paragraph. 

https://www.ieee802.org/3/dj/public/adhoc/optics/0823_OPTX/ran_3dj_optx_01_230815.pdf

https://www.ieee802.org/3/dj/public/adhoc/electrical/23_0817/ran_3dj_01_230817.pdf

Chris

On Wed, Nov 15, 2023 at 6:54 PM Chris Cole <chris@xxxxxxxxxxxxxxx> wrote:
It was nice to see the "Update on Consensus for 500m and 2km Optical Objectives", lusted_3dJ_07_2311, and the dropping of the awful proposals for 250m objectives using AI/ML as an opportunistic justification since there are no meaningful technical differences with respect to FEC between 250m and 500m.

It is a bit unfortunate that we rushed the adoption of the 500m FECo baselines because the technical feasibility has not been shown, in particular we have not seen waterfall curves which means we don't know the low error floor range and the receiver sensitivity. 500m FECo is the right place to end up, but right now the baseline values are not based on measurements. Hopefully these will be brought in even though the motivation to do so is less.

Since we adopted 500m FECo, this puts a hard constraint on C2M specification. Any C2M proposal that does not have significant implementation margin, needs to demonstrate that the end-to-end link works. Brian's presentation touches on some of the limitations of just using pre-FEC BER (https://www.ieee802.org/3/dj/public/23_11/welch_3dj_02a_2311.pdf). An observation is that FLR is important which is not cleanly partionable when using FECo optics if there is no C2M margin.

Chris

To unsubscribe from the STDS-802-3-B400G list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-B400G&A=1