| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Colleagues, As I noted in the July meeting, I intend to host a couple of meetings to discuss required changes to jitter specifications. If you are interested in joining please reply to
adee.ran@xxxxxxxxx to receive invitations. My default schedule is Wednesday morning PST. If you intend to join, please let me know if this is a show stopper for you. Rough agenda: As some people mentioned during the July meeting, the jitter specifications used in clause 72 and its ba descendants are incomplete and possibly inadequate for the equalized copper channels. Current draft still has TBD notes for both values and methods (clauses 92.8.3, 93.8.1.7 and 94.3.11.9, and also tables 92-8, 93–7 and 94-7 which are related). We should start building consensus on a proposal. To address the new challenges of 100G Ethernet, I think we should improve the discrimination between the multiple causes of the phenomena we call "jitter". Some of these may be much more tolerable than others. Some might be better counted
in other places. Putting many of them in the same bucket is technically incorrect. Measurement methods should be adjusted to what we actually intend to measure, and compliance values should be made both realistic and linked to expected performance. I intend to limit the scope of this discussion to TX specifications (clauses 92.8.3, 93.8.1.7 and 94.3.11.9) and only briefly touch RX tolerance implications. Regards, Adee Ran |