Sector, Terence, and All,
This information surrounding SilkRoad is interesting, and is causing a lot of speculation both market-wise and technologically. See other comments on the Ciena and Gilder threads and elsewhere during the past three days.
I wouldn't be too taken about the single fiber strand aspect, however, because most responsible telecomm managers always want a back door to protect against route and element failures (ol' man back-hoe). Most architectures that permit restorative capabilities through seamless self-healing techniques usually are based on at least two diverse and redundant physical layer paths.
This should not suggest that redundant paths couldn't be deployed when using the Fiber Driver, or with the proposed SR product fot that matter, since two of anything, in the end, can be procured if there are enough port and address accommodations.
It may suggest, however, that if two physical paths are indeed needed to satisfy these diligence-related measures (and they usually are by the largest of customers, when there are very-high capacity links at stake), then this would only serve to negate the single fiber advantage, while possibly adding to the cost of terminal gear and administration.
Just my observations as to the relative merits of a single strand approach, per se, for whatever it's worth.
Regards, Frank Coluccio |
