Time Warner Cable Increases Workforce in Kansas City > Competition

Competition

I do not think TWC will tip their hand until they see that Google Fiber is actually providing service to the fiberhoods that qualified for service. Google is famous for project delay and abandonment. If and when Google Fiber actually provides dependable service to subscribers, TWC could put out a 300Mbps download/100Mbps upload offering with a similar TV package at a lower price than Google Fiber+TV. The upgraded internet portion is minor in cost increase for them. 8 x 4 channel bonding modems already exist. What I see as a potential problem is the possibility that content producers will make deals with Google that keep the TV package and TV add ons at a lower customer cost then they have with TWC. Reason would be to be part of new advertising models and extremely low signal compression issues. Imagine 8 1920x1080P video feeds with no compression and 1Gbps internet running at the same time. TWC cannot provide that kind of platform yet for content delivery and interaction with the viewer.

TWC would have to do some serious upgrades to be able to handle a 300Mbps tier in any great numbers, as in actually trying to come close to competing with Google fiber and still only providing a fraction of the bandwidth. The 100Mbps up would be the hardest part. DOCSIS to date was not designed for having so much upstream speed and trying to do so with a lot of people on each node. The next DOCSIS update will be focusing heavily on the upstream situation and working towards greatly increasing the bandwidth available.

The DOCSIS 3.0 speed tables, »en.wikipedia.org/wiki/Docsis#Speed_tables, show that upstream is any number of bonded channels multiplied by a net 27Mbps per channel. The present D3 modems have 4 upstream channels, which should give 108 Mbps if all 4 channels are bonded. Intel's new Puma 6 may allow upstream bonding of 8 channels resulting in 216 Mbps upload and downstream bonding of 24 channels for 912 Mbps download, resulting in a 4.22 down to up ratio. Not the same as 1Gbps symmetrical, but very good if the same coaxial cable can be used. There are rumors that 32 bonded channels down, resulting in 1216 Mbps download, will be achieved in the next Puma. Unfortunately, it seems that upstream may be stuck at 8 bonded channels, so you would get a 5.63 to 1 downstream to upstream ratio.

Yes, so you can have a handful of users at the most on each node of 500 - 1000 homes with 1Gbps speeds and that's it. That's not even coming close to competing with Google fiber.

I do not dispute the need for reducing the number of subscribers per node by splitting nodes or by upgrading the fiber from the node. I was showing that the modem and CMTS manufacturers have done their part and will do more in the future. The ISPs have to do their part to make it truly work. Hybrid fiber coaxial will never work as well as pure FTTH, but many communities would welcome the upgrades to a better version of hybrid fiber coax in the future if there is no other choice. Since the analog mandate for cable television has ended, it would be nice to see expanded efforts to use some of the cable plant efficiency gained to go into improving internet performance.