Before your comments get too far out of hand I better get things back somewhat on track:
Broadband is a medium to a branch of the backbone. It does not threaten the backbone. In fact, the more bb proliferates, the more backbone bandwidth WCOM will be able to provide.
The first sentence is false.
The second sentence is incoherent.
The third sentence begs itself.
ATHM does not have much content, so content providers are connected to ATHM ISP just as other ISPs are connected to those providers.
Physically the connection is quite different. You should study @Network.
One could infer from your above statement that ISPs don't provide content or that ATHM has no more content than do copper ISPs. However, the quality of access enables the superior content whomever is the originator. ATHM can get just about anything a copper ISP can, but the opposite isn't true. This is the biggest reason T bought TCI.
So, as ATHM relieves the bottlneck between user and the ISP (ATHM) through broadband, the congestion on the backbone may become the bottleneck.
90% of copper ISP trouble comes from backbone overload or breakdown. ATHM leases a private 'bone, actually, private leased lines from a carrier I won't mention and so is insulated to some extent from heavy traffic problems like latency commonly experienced with copper ISPs. The private 'bone is not physically disjoint from other carryings and so ATHM can be crowded or crowded out depending upon ATHM level of QoS and the volume of other priority feeds. Currently 'bone capacity is adequate, but there are transients when that isn't the case. If another carrier's bone drops and alternative lines are loaded, the entire system to which ATHM is only partially connected, is impacted. Thus, ATHM is vulnerable too at the 'bone level. Most of the local loop problems have been caused by local origination, not by 'bone problems.
There will also become more demand for caching by content providers and ISPs, as one way to reduce contention by getting content closer to
the end user.
The only way this makes sense is that by caching you mean compression. Caching takes place on caching servers in or near the local loop. That is quite removed from the providers who upload their content through NAPs to the @Network. The purpose of caching is to relieve the download side burden by shunting a local loop request to a cache to avoid superfluous loading of the edge. ATHM is the only ISP which has this redundant parallel ATM network interacting through ethernet driven HFC loops.
|
