|Thread Links||Date Links|
|Thread Prev||Thread Next||Thread Index||Date Prev||Date Next||Date Index|
While we are on the subject of "service to humanity..."
I think that it will benefit everyone to define Clause 45 functionality to cover all xxBASE-T PHYs. It will remain optional to use either Clause 22 or Clause 45 access but it will allow PHY vendors the opportunity to migrate to the Clause 45 for all of their newer devices. This has the advantage of the lower signal levels as well as creating space for support of a wide range of "new ideas" such as line testing. It means that (eventually) system vendors will be able to use a single management interface for xxx + 10 devices (just like the good old days, before 10G). It will also be compatible with EFM copper PHYs - which is clearly a "good thing."
Don't forget that we have also defined Clause 22 access to Clause 45 for those that want backwards compatibility with existing management stations.
"Booth, Bradley" wrote:
Those are the questions to be answered. I think there is a possibility of repeating what we did in 802.3ae. One goal that we tried to shoot for was to have the register sets of Clause 22 and 45 look very similar, so that ties in with your idea of having the Clause 22 functionality move into Clause 45 as a xxBASE-T PCS. There will be a need for 10GBASE-T registers in Clause 45 to have similar functionality to those found in Clause 22. If 10GBASE-T uses the similar functionality and register set as that in Clause 22, mapped into Clause 45, then we should be able to use a xxBASE-T PCS status register that can indicate the PHYs abilities from 10, 100, 1G and 10G. I think as we move forward devices will start to support Clause 45 MDIO (especially those designed for 10G or faster) but will still offer legacy support for older PHYs (migration path). The key will be to enable the legacy xxBASE-T speeds built into 10GBASE-T PHYs to be accessed via one Clause 45 interface, without the need for a Clause 22 interface. A sort of "service to humanity" thing. :-)