[STDS-802-3-25G] 答复: [STDS-802-3-25G] some question about the current D1.1, can any one help me to understand it right

[Zhongqiwen <zhongqiwen@xxxxxxxxxx>]

Arthur

 

Thanks for your clarification : ), it’s clearly for me now : )

 

Best regards

Qiwen

 

 

发件人: Arthur Marris [mailto:arthurm@xxxxxxxxxxx]
发送时间: 2015年7月10日 16:08
收件人: Zhongqiwen; STDS-802-3-25G@xxxxxxxxxxxxxxxxx
主题: RE: [STDS-802-3-25G] some question about the current D1.1, can any one help me to understand it right

 

Qiwen,

    Clause 74 FEC is only mandatory for the four copper twin-ax and backplane PHYs. Selection between Clause 108 and Clause 74 FEC for the 25GBASE-CR and 25GBASE-KR PHYs is done by the auto-negotiation function (described in Clause 73).

 

    The MMF 25GBASE-SR PHY only uses Clause 108 FEC (never Clause 74) so there is no need for a selection mechanism for the 25GBASE-SR PHY.

 

    The 25GMII and 25G PCS are a simple re-use of the existing 10G XGMII and PCS and the 25G RS-FEC is not used by all the 25G PHYs. In order to allow re-use of existing 10G implementations for the 25GBASE-CR-S and 25GBASE-KR-S PHYs it was decided to move the idle and delete functions for codeword markers to Clause 108 as the RS-FEC is the only sub-layer that requires this functionality.

 

   I hope this helps.

 

Arthur

 

From: Zhongqiwen [mailto:zhongqiwen@xxxxxxxxxx]
Sent: 10 July 2015 06:32
To: STDS-802-3-25G@xxxxxxxxxxxxxxxxx
Subject: [STDS-802-3-25G] some question about the current D1.1, can any one help me to understand it right

 

Hi, everyone:

 

1,  in table 105-2, the clause 74  are mark “M” BASE-R FEC for the 5 PMDs,  and Clause 108(~Clause91) are Mark “M” for CR, KR, SR,

     My question is:  are those 2 FEC options need to be select one for use? Or …???

 

2, the current 25GMII and PCS are scale up from 10GE, while using RS-FEC, AMs need to be inserted. The rooms for AMs are ready in 100GE for 802.3bj clause 91 FEC(remap), but not ready for 25GE that scale up from 10GE PCS.

So the Figure 108-2 shows a rate compensation for making rooms for codeword markers. This implying Idle insert and delete operations. And thus imply descrambling, (decode), Idle insert/delete, (encode) descrambling.

 

My understanding is, this functionality of making rooms for AMs were done in PCS over CGMII/XLGMII, we should move this functionality to the Clause 107, as one of the optional enhancements to clause 49 PCS(simple text described 107.1.2 consist with 802.3ba, rather than to  show it in figure 108-2, we can refer to Figure 82-2) while using RS-FEC.  

 

Is my understanding right ? or there are other considerations?

 

Best regards

 

Qiwen