[opendtv] Gigablast

From: Craig Birkmaier <brewmastercraig@xxxxxxxxxx>
To: OpenDTV Mail List <opendtv@xxxxxxxxxxxxx>
Date: Sat, 07 Nov 2015 11:59:28 -0500

I'm starting to see commercials on Cox Cable for their new Gigablast service.
Going to their website they give the impression that it will be available
across the Cox footprint, however, the reality is that they are deploying a
mixture of technologies with fiber to the home in select markets, primarily in
NW neighborhoods.

This article adds some light...

http://www.multichannel.com/news/technology/cox-expands-gigabit-service/390403

Using fiber-to-the-home technologies in a targeted fashion, Cox first
launched the G1GABLAST in Phoenix last fall.

Cox also expects to factor in DOCSIS 3.1, an emerging multi-gigabit platform
for widely deployed HFC networks, as part its grander plan to make
residential gigabit service available across Cox’s footprint by the end of
2016.

The FCC now defines broadband as 25 Mbps downstream / 3 Mbps upstream. I've run
some numbers against this for DOCSIS 3.1, assuming everyone getting broadband
from a cable system PON would need between 25 and 50 Mbps to move from the
current broadcast MPEG-2 video infrastructure to an all IP streaming
infrastructure.

But streaming may be the wrong way to look at this...

The Cox website tells us:

Download 100 tunes in 3 seconds, an HD movie in under a minute and more.
Except for live programming, most OTT streaming is actually done using
progressive download to local cache. You can see this on the timeline/progress
bar in most media players.

This raises the question of how provisioning video services to a neighborhood
may actually be done, and what the real bandwidth requirements are if it is
possible to "blast" a program to local cache.

Obviously anything that is taking place live cannot be downloaded in advance;
live programming will look much like the MPEG-2 TS streams that exist today. It
is less clear whether linear networks will treat everything as "live
broadcasts" or move to a download business model (or do both).

So the question I have is: does the ability to blast a program to local storage
change the requirements for total bandwidth to a node of a HFC cable system?

And the follow-up question.

If gigabit speeds are just around the corner does any of this really matter?
Will broadband infrastructure blast past the "video problem" by the end of this
decade?

Regards
Craig

[opendtv] Gigablast

Other related posts: