Ken,
I think that Michael Armstrong and Co are aware of this. Or, at least they have been advised. We don't want to skew Mike's thoughts here with too many facts while he's still on a run.
That's why we've seen their engineers making some rather aggressive projections concerning how they are going to remedy the situation. Three to five years out may not seem aggressive, but given the scope of this thing, it's a lightning strike.
There are some stats available, but I don't think that they yield significant meaning yet. Shared domain networks will scale linearly from zero up through some percentage of utilization, and then suddenly go flat for the next couple of increments of traffic increases, and then begin their decline with added users pulling down incrementally larger content payload sizes.
These dynamics aren't entirely what one would perceive to be the case, nor can the knee in the curve be easily predicted scientifically, much less intuitively, is what I'm saying.
The answers to some of these questions can be found in the results of the stress tests which (one would hope) were taken during the initial and ongoing HFC trials.
If those stress tests were performed with any foresight and diligence, very heavy continuous streams would have been included, along with those for "typical" flows with increasing usership. I don't know if the present anticipated loads were even anticipated then, though, so I really can't say for certain that the early trials even took them into account. Don't know. Anyone have the inside on these?
By the way, and I need to insert this here... the last couple of posts of mine are predicated on the assumption that uptake levels will be achieved to make these predictions come true. There is no guarantee that that will happen. Early adopters aside, I know plenty of folks who are quite content right now with other modes of access, and competitive forms have yet to be unveiled. So the saving grace for HFC, two or three years out, may be its failure to become overly attractive. Hence, the same paradox as earlier cited, from another angle.
Regards, Frank Coluccio |
