Jack, I don't think that your assessment is entirely valid in principle, although the boxes will surely go, as part of what I think you are implying. See my previous reply to Kenneth in the upstream. In your post you noted:
"TDM stuff. Hardly worthy of note as it is rapidly becoming obsolete."
CIEN and others will do the same things that have always been done (see clarifications below) with regards to SONET add-drop, grooming, etc., until lightspeed switching at the photonic level is possible and economically viable, and until end users and carriers alike cease to use those formats to define their units of currency. E.g., OC-3, OC-48 and OC-192. Even the STS-1 level (equivalent to what would be an OC-1) at the electrical level is highly traded these days in colos and hotels, and the photonic box makers are accommodating these, too.
SONET and SDH framing, in particular, will withstand the forces of evolution for quite some time, until another unit of currency is minted out of the standards bodies that everyone can agree on. Witness how even Ethernet is now being subsumed into these standards at the 10 and future 40 Gb/s rates, although there will be more than one way to haul Ethernet, to be sure.
Of course, they will also be doing other things, like trans-framing from one format (i.e., SONET) to another (Gbe), and so on. These processes, too, have a history in using rows and rows of discreet, legacy boxes, which are the "real" targets, for their clumsiness, that you are addressing in your post.
The clarifications I alluded to earlier: The main difference here being that the newer, primarily-optical boxes are driving these processes deeper into their own backend switching fabrics, as opposed to using discreet boxes to achieve the same end. Thus, hiding all of those ugly and expensive (to buy, house and administer) boxes from view. The CD, as a case in point (along with a growing number of other vendors' wares), does some of this now and will be doing more of the same, in addition to the more-obvious lambda-centric, a la MPLS and MPLamdaS switching and routing, schemes, too.
Corrections and further comments are always welcome.
FAC |
