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

Re: [STDS-802-3-400G] 400Gb/s Electrical Interface Thread - Simulating Coding/Modulation over Discussed Channels

Mike/Joel 

I do agree with Mike that option 5 is a very valid one.

It is good to have a table with all possibly options, but the focus need to be on the MR and VSR as these two interfaces only had broad support during Sept 802.3bs meeting.

All optical interface not necessary have to have the same signaling, since VSR has a PMA-PMA chip.  On the other hand XSR and USR will not have a PMA-PMA so in my opinion they don’t need to have a same interface as VSR.

Thanks,
Ali Ghiasi
Ghiasi Quantum LLC
Office (408)352-5346

On Oct 8, 2014, at 9:49 AM, Mike Dudek <mike.dudek@xxxxxxxxxx> wrote:

I think you have missed the
 
Then (5) modulation method is the same over LR and MR, and the same different modulation method is used for VSR, XSR, and USR.  (ie all optical module interfaces)
 
Mike Dudek 
QLogic Corporation
Director Signal Integrity
26650 Aliso Viejo Parkway
Aliso Viejo  CA 92656
949 389 6269 - office.
 
 
From: Joel Goergen (jgoergen) [mailto:jgoergen@xxxxxxxxx] 
Sent: Wednesday, October 08, 2014 9:32 AM
To: STDS-802-3-400G@xxxxxxxxxxxxxxxxx
Subject: [STDS-802-3-400G] 400Gb/s Electrical Interface Thread - Simulating Coding/Modulation over Discussed Channels
 
All
 
We have been working towards guidelines/limit lines of channels appropriate for 50Gb/s electrical interfaces within the 400Gb/s physical system architecture.  There will be discussions on that in the adhoc meeting tomorrow morning.
 
But that is still a few steps away form working up proposals of modulation methods over an 8 lane by 50Gb/s interconnect.
 
We have talked about potential reach categories, potential reach budgets, and there has been a number of channels uploaded at
 
From a modulation stand point, we need to put a proposal together in the coming weeks.
If you or a colleague have time to look at modulation concepts over a range of channels in the next week, and would be willing to share that in 2 weeks towards a consensus proposal, that would be great!  
 
I would like to start some discussion on this …
 
It seems to me that we could …
Given C2F (LR), C2C (MR), C2M (VSR), XSR, USR
Then (1) modulation method is the same over LR, MR, VSR, XSR, USR
Then (2) modulation method is the same over LR, MR, VSR and not XSR, USR
Then (3) modulation method is the same over MR, VSR, XSR, USR and not LR
Then (4) modulation method is the same over MR, VSR and not XSR, USR, LR
 
(1) would be the perfect world but I honestly don't think it will physically work well in applications that make sense to architect.
(2) and (3) seem to be reasonable.  There are advantages to (2) with potential receiver compatibility across LR, MR, and VSR.
 
<image001.png>
 
 
 
Take care
Joel Goergen