Blog Post

Will TV Everywhere Swamp Cable Networks?

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

[qi:032] Comcast (s cmcsa) may take on more than its network can handle by offering its cable TV via the web under its TV Everywhere program, which has me wondering if cable providers will weather the influx of TV content delivered over their data network as opposed to their video network. Unsurprisingly, Comcast and industry equipment vendors say the network will be just fine for TV Everywhere delivered as so-called over-the-top video.

But streaming potentially popular cable TV content over cable companies’ limited pipes could cause problems for the providers. Cable companies share the last-mile network among a certain number of homes, which is why one person uploading P2P files can seriously degrade the connection for other users. Plus, the companies usually allocate only one 6 MHz channel of their spectrum for data services, which translates to about 31Mbps throughput.

If you use a rough estimate of 2Mbps to 3Mbps as the amount of throughput needed for a single SD video stream and assume that 20 households in a 500-home node are trying to watch the stream, that would take up to 60Mbps of capacity. But that 60Mbps is now more than what is available in one DOCSIS channel, which means cable providers will have to allocate more spectrum resources for data. Cable providers maintain that this is all hypothetical and that anyone hoping to break the cable network would have to radically change their viewing habits. Charlie Douglas, a Comcast spokesman, says that the company isn’t concerned about the effect TV Everywhere will have on the network, and pointed me to a Nielsen report that showed that people watch only 3 hours of online video a month vs. 153 hours on their TVs.

On the technical side, a Motorola (s MOT) executive said that cable companies have several tools to make sure they spool out enough bandwidth resources to support a program like TV Everywhere. Cable companies can split the node, which essentially halves the number of homes that the equipment nearest the homes will serve, they can switch analog channels to digital ones, which are more efficient users of spectrum, and they can bond channels together if the provider introduces DOCSIS 3.0 to make faster pipes.

Floyd Wagoner, a director of marketing and communications for Motorola Access Networks Solutions, said these tools will allow the cable guys to meet demand for broadband at least for the next decade before they have to move to all-fiber networks. Most cable companies already have fiber in the ground to the head-end equipment, which helps them support efforts to deliver services that use more bandwidth. Motorola sells equipment to help cable companies implement these bandwidth-boosting steps.

Wagoner also notes that cable providers are vigilant watchers of their networks, and react swiftly as demand for broadband rises. Comcast’s Douglas declined to share the company’s forecast on how much bandwidth TV Everywhere would consume, but it’s safe to say it has predictions about how it will affect the network. As to how it will affect your bandwidth cap, well that’s a different story.

Bonus Link: An awesome bandwidth calculator on the Motorola web site lets you plug in estimates about the number of homes served, cable penetration (average is about 30 percent), and what type of speeds and concurrent use models are supported for data or streaming. In general, high-speed data averages a 2 percent concurrency and video on demand (which is closer to a streaming model) estimates a 10 percent concurrency.

5 Responses to “Will TV Everywhere Swamp Cable Networks?”

  1. All Comcast has to do to ensure that TV Everywhere doesn’t swamp their pipes is cache the stuff at all the appropriate places, and they’re clearly smart enough to do that.

    Most of the Comcast engineers I know are actually pretty good at arithmetic, as strange as that may seem.

  2. Hmmm..will comcast overburden their own pipes following some IPTV pipe dream? Not likely. Technical underpinning apart, this article fails to inspect the business logic of comcast’s move. Comcast can no longer be faulted for not offering an on-demand service. Comcast can try to get some money for content they now store in their shelves – content that is potentially monetizable, but they cannot because their video pipe is filled with other (more monetizable) content. Comcast can now gain some brownie points against the hulus of the world. Ofcourse this business logic can fall flat if all of comcast’s customers sign up for this service and all of them switch to seeing TV Everywhere at the same time. Comcast will ofcourse guard that this will not happen, by making sure that the content on this service is not as compelling as the one on their video pipe – for example, they wont screen live sports events on this service. Nor will this stream contain american idol or contemporary news items. So, if you want to watch a 90s oscar winning movie, you could save yourself a trip to the DVD rental (or a subscription to netflix/roku) and watch it on TV everywhere. If you want to watch the game, you just watch plain old tv.

  3. This article is not well thought out. Tv everywhere is no different than Hulu and will just add more content. It will be 2 years before all of the industry and all of the content is available and even when it is no one is going to stream to a PC at rates much above 1.5 mbps. Even if 50% more people start using it , it will actually spread the bandwidth around between cable/ telcio sites like comcast and TWC and AT&T and between content guys like Turner and Discovery. A large share will be sure to go to Hulu and all of the people downloading bit torrent aren’t going anywhere.

  4. Two errors in this posting. First, “Internet” video isn’t typically encoded and delivered as
    at 2-3 mbps (mpeg-2 is typical for most cable providers). It is more likely the delivery system
    would use a more modern codec designed for PC use, or H.264. In either case, it would
    be reasonable to expect the delivery bit rate to be closer to 1 mbps. For example, 700 kbps
    looks like fine on an iTouch.

    Then in the Motorola link, that also assumes the cable industry standard of
    3.75 mbps for SD video on demand.