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

Re: [802.3_100G-OPTX] 802.3CU ER and OMA specs



Hi Brian,

 

Thank you for initiating a reflector discussion of this spec.

 

There has always been a trade-off in standards between the desires of users and suppliers. The users want maximum performance for nothing, and the suppliers want to at least eke out a living. Brian is bringing the supplier perspective.

 

However, Brian skillfully avoids confronting user requirements, by framing this as a manufacturing margin vs. documentation minutia debate. We mostly don’t printout standards anymore, so saving on ink by having fewer lines in a spec. is not a major consideration.

 

Below is a complex diagram showing how network engineers view what optics technology does for them:

 

A____________________________________________B

 

They don’t care about such things as ER, OMA, TDECQ, or mathematical relations between them. This comes as a shock to us because our life is wrapped up in this stuff, and we all want to be appreciated.

 

Users just want to know what budget they get to go from A to B. It used to be that power budgets didn’t matter to the user. They only cared about channel insertion loss. But that’s no longer the case. Optics in the data center are used at much shorter reaches than maximum, so RS is as important, if not more important than SRS.

 

This is where a standard that can’t tell the end user what power budget they get, fails. Providing a menu of choices as function of ER is not being clever, it’s not doing our job.

 

If 0.1dB manufacturing margin is important today, we can consider using the power budget for ER > 4.5dB. When in the future, we get around to DMLs for this application, we can worry about their manufacturing margin.

 

Whatever we do, let’s at a minimum state what power budget is supported by the standard.

 

Thank you

 

Chris

 

From: Brian Welch (bpwelch) <00000e3f3facf699-dmarc-request@xxxxxxxxxxxxxxxxx>
Sent: Friday, May 22, 2020 11:33 AM
To: STDS-802-3-100G-OPTX@xxxxxxxxxxxxxxxxx
Subject: [802.3_100G-OPTX] 802.3CU ER and OMA specs

 

Mark/All,

I won’t have a presentation for Tuesday, but I did want to follow up on a discussion that was had in our prior meeting. In the presentation “Nicholl_3cu_051920” there was documented an attempt to reconcile several comments, including to simplify specifications are it regards different extinction ratio ranges (“high” vs “low” ER). One thing that first became apparent to me during this meeting was that the proposal to do this resulted in an increase in the power budgets for the 100G/L specs of 0.1dB (for high ER), and subsequently an increase in OMA specifications by the same amount.

 

While I’m not opposed to simplifying the spec, I would prefer to do it in a way that doesn’t end up also making it more stringent. While 0.1dB may seem like a trivial amount the effect of the change can be real, especially in how moving a spec limit will degrade the manufacturing statistics of a part (reduced Cpk, which could drive more stringent testing requirements, amount other things). The optics world today is one where gains are often found a fraction of a dB at a time, so to sacrifice some of those gains when it is not absolutely necessary seems unfortunate to me.

 

As such my preference for now would be to keep the transmitter specs (and power budgets) as they are, where we report “OMA-TDECQ” for two extinction ratio ranges (consistent with how it is specified for 100GBASE-DR and 400GBASE-DR4).

 

I would be interested in others thoughts on this matter.

 

Thanks,

Brian

 

From: Mark Nowell (mnowell) <00000b59be7040a9-dmarc-request@xxxxxxxxxxxxxxxxx>
Sent: Thursday, May 21, 2020 6:30 AM
To:
STDS-802-3-100G-OPTX@xxxxxxxxxxxxxxxxx
Subject: [802.3_100G-OPTX] Reminder on presentations: 5/26 P802.3cu Interim Teleconference

 

Dear Colleagues,

 

This is a reminder about  our deadline for presentation requests for next week’s P802.3cu Interim teleconference.

 

Please send me your presentation requests by Thursday May 21st (today) and send me the presentation by Friday May 22nd (tomorrow).  A reminder that with the limited time we have for these teleconferences, presentation time will be limited to no more than 15 mins.

   

Please note that All interim teleconference meeting participants should review the following documents prior to participation in an interim meeting teleconference: 

    • IEEE SA patent policy
    • IEEE SA Copyright Policy
    • IEEE SA Participation Policy

 

All of these policies may be found at http://ieee802.org/3/policies.html.

 

Regards,  Mark  

 


To unsubscribe from the STDS-802-3-100G-OPTX list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-100G-OPTX&A=1


To unsubscribe from the STDS-802-3-100G-OPTX list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-100G-OPTX&A=1


To unsubscribe from the STDS-802-3-100G-OPTX list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-100G-OPTX&A=1