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

Re: [802.3_B400G_OPTX] Historical Perspective on Transmitter Penalty

The second sentence below should say:

"The issue in 802.3bs was that a functional link was difficult to define because there was almost no hardware available to take measurements, ... 

Chris

From: Chris Cole <chris.cole@xxxxxxxxxxxx>
Sent: Tuesday, July 29, 2025 8:33 AM
To: STDS-802-3-B400G-OPTX@xxxxxxxxxxxxxxxxx <STDS-802-3-B400G-OPTX@xxxxxxxxxxxxxxxxx>
Subject: [EXTERNAL]: Re: [802.3_B400G_OPTX] Historical Perspective on Transmitter Penalty
 
Ali

Thank you for the bringing this to our attention.

The issue in 802.3bs was that a functional link was not difficult to define because there was almost no hardware available to take measurements, like presented by Roberto this AM, and like we will see a lots more if the FRx/FEC proposal is added to the draft. The irony is that there was no hardware and no measurements in support of TDECQ to show that it's not well correlated to actual link performance. 

We got away with just TECQ/TDECQ for 50G/lane and 100G/lane optics because there was margin. Evern at 100G/lane it has been problematic as noted in the FRx/FEC proposal presentation. This has contributed to undermining confidence in 802.3 optics specifications. 

We can't go back and change what we did in 802.3bs, but we do control what we do in 802.3dj and we should do everything possible to add the functional link test to complement TECQ/TDECQ. This is also the best way for TECQ/TDECQ to have a chance at usefulness. 

Thank you for reminding us that Jonathan was an industry giant whose contributions were consistent over decades and had a profound and lasting impact on our industry. It's brings a sense of pride, as a well as sadness about his untimely passing.

Chris
From: Ali Ghiasi <aghiasi@xxxxxxxxx>
Sent: Tuesday, July 29, 2025 3:06 AM
To: STDS-802-3-B400G-OPTX@xxxxxxxxxxxxxxxxx <STDS-802-3-B400G-OPTX@xxxxxxxxxxxxxxxxx>
Subject: [EXTERNAL]: [802.3_B400G_OPTX] Historical Perspective on Transmitter Penalty
 
 
Hello All,

Below is the JK contribution from 2015 that I referenced during Ahmad presentation Q&A.  Below is statement from JK presentation from 2015!
What Jonathan says in his presentation is exactly the issue we are debating in 802.3dj that TDECQ is necessary but not sufficient to capture long pattern effects!
According to Jonathan TX quality metric is most repeatable in software but hardware RX “Functional Receiver” is best to capture long pattern effect and block BER.
Jonathan presentation pre-date TDECQ definition and provided strong justification to go ahead and develop TDECQ.



Tx quality metric should include Ref Rx and Ref EQ

• most repeatable in software

• hardware Rx and Eq capture long pattern effects and allows a 'real'

BER test.

4


preview.png

Thanks,
Thanks,




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

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

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