Your last several posts on gluts add perspective to the claim of bandwidth glut. On the surface, the massive amount of fiber (and associated empty conduits and other spaces and pathways) provisioned in the ground gives the illusion of an oversupply of 'bandwidth.'
In reality, these dormant strands represent a very low cost component of latent capacity, i.e., potential or a form of buffer stock as you mention in your next post;-) just sitting there waiting to be activated at the right time. It's only at the point of provisioning, when the optoelectronics and span amplifiers and repeaters are installed and activated on those strands, that the claim of bandwidth existing can be made.
This last level of provisioning is what costs the big bucks, comparatively speaking, and not laying the additional fibers, themselves. Adding the extra strands, in reality, simply enhances the economic outcome of those builds due to factors having to do with scale.
Having said that, in many of the venues that one would want to consider (your petrol plants and aeroplanes, for example), similar levels 'latent' product potential cannot be as easily converted to real product as quickly. Perhaps smart power techniques will change this in the electric utility space. But it's hard to envisage just in time delivery of a Boeing 777, or an AirBus A3xx, unless, as you say, they are retreived from moth-ball status and reactivated.
It's for this reason that perhaps some carriers' strategies to oversupply the raw materials (placing additional silica in the ground) in advance is more than simply justified, because having to start laying the low-cost-component- fiber anew, from scratch, each time additional bandwidth was required would otherwise take several years to have accomplished - several years of extremely capex intense activity, including labor-intense trenching and other construction costs - as opposed to simply firing up what already is in the ground with a couple of lasers and wdms.
This is not to suggest that every carrier need do this. Even some of the larger carriers, whose primary products are vertically integrated services, say, might be best served on certain routes by smart buying capacity from others (IRUs, spot purchases, whatever), as opposed to undertaking their own construction efforts in many cases. In the carrier world, as elsewhere, it's usually smart buying that matters, even more importantly than smart selling.
These may be gross oversimplifications of the actual processes and considerations that take place in the greater scheme of capacity planning and management, but I think I make my point.
FAC |
