Re: [RPRWG] RPRWG[PHY]

["Robert D. Love" <rdlove@xxxxxxxxx>]

Title: RPRWG[PHY]

Thank you Harry.  Though your review may not be exhaustive, it is a good start at taking our jobs seriously and carefully looking at the implications of what we are specifying.  I appreciate your efforts.

 

Best regards,

 

Robert D. Love
President, LAN Connect Consultants
7105 Leveret Circle     Raleigh, NC 27615
Phone: 919 848-6773       Mobile: 919 810-7816
email: rdlove@xxxxxxxx          Fax: 208 978-1187

----- Original Message -----

From: Harry Peng

To: RPRWG

Sent: Wednesday, March 19, 2003 11:11 AM

Subject: [RPRWG] RPRWG[PHY]

As Geoff pointed out, we still have to do our homework.
Here is the start. I took a quick look at the 802.3PHY sections
Here are some specification with exposed interfaces:

24.6.1 PHY delay constraints (exposed MII)
Every 100BASE-X PHY with an exposed MII shall comply with the bit delay constraints specified in
Table 24-2. These figures apply for all 100BASE-X PMDs.

Since we are not doing 100M so this clause does not apply....

35.2.4 MAC delay constraints (with GMII)
A Gigabit Ethernet MAC with a GMII shall comply with the delay constraints in Table 35-5.

The only thing of interest here is with COL declaration, which does not apply to 802.17

The other issue is jitter. Here is where packet based transmission is simpler because
the synchronization is only required for a duration of an MTU. I will do some calculation
as to how many UI of  jitter we can tolerate with +/- 100ppm clock and still can handle jumbo frames.

.... There are also the 10G PHY's.......

Regards,

Harry