http://www.businessweek.com/news/2010-02-10/google-plans-to-build-high-speed-fiber-optic-networks-update2-.html
http://googleblog.blogspot.com/2010/02/think-big-with-gig-our-experimental.html
What do folks think?
Granted it's very early on, and g00g could decide to never leave the
announce phase.
- --
Charles N Wyble
Linux Systems Engineer
(818)280-7059 charles@knownelement.com
Optimistic view: It can force the incumbents into being competitive on
service and everyone wins.
Pessimistic view: incumbents feel threatened and try to sue/lobby it
away to keep the status quo like they did with cities trying to offer
wifi or FTTH.
~Seth
I think it's great!
I've been preparing to float a similar idea locally.
If this is how they use their market cap, I would love for them to do it in my local market, which does seem to hold a near-and-dear place in the heart of some google C* types.
- Jared
* Local details/breakdown: http://puck.nether.net/~jared/blog/?p=84
>
http://www.businessweek.com/news/2010-02-10/google-plans-to-build-high-speed-fiber-optic-networks-update2-.html
>
http://googleblog.blogspot.com/2010/02/think-big-with-gig-our-experimental.html
>
> What do folks think?
>Optimistic view: It can force the incumbents into being competitive on
service and everyone wins.Pessimistic view: incumbents feel threatened and try to sue/lobby it
away to keep the status quo like they did with cities trying to offer
wifi or FTTH.
Google cash > Muni cash. I'm not saying it'll work, but they have many more
resources at their disposal. Incumbents should be worried.
I'm really interested in their distribution ideas, as well as the
bottleneck from the Google network to the rest of the internet.
Ah, who am I kidding, it's not like anyone cares about the rest of the
internet, right?
--Matt
Jared Mauch wrote:
I think it's great!
I've been preparing to float a similar idea locally.
If this is how they use their market cap, I would love for them to do it in my local market, which does seem to hold a near-and-dear place in the heart of some google C* types.
- Jared
* Local details/breakdown: http://puck.nether.net/~jared/blog/?p=84
Awesome write up.
Has anyone in the NANOG community been approached by google? I mean
presumably this would require a massive coordination effort with
existing exchange points etc. Or is google going to simply build an
entire long haul network as well? Perhaps combine this with the containers?
My assumption was Google already had their long haul network up and running
(since 07-08):
http://www.google.com/search?q=google+dark+fiber
-brandon
The WSJ says: "In an interview, Google product manager Minnie Ingersoll said consumers
will be able to buy service directly from Google or from other
providers, whom Google will allow to resell the service. She said
Google will manage the deployment of the network but probably partner
with contractors to help build it."
--Steve Bellovin, http://www.cs.columbia.edu/~smb
Thanks. I want to codify it to something more (average) human-readable before I socialize it in the local community.
This sort of investment could have some immediate payback, esp if you have local utility (water, power) buy-in. The challenge I see is having the political will to undertake the project. If you adjust rates up over the first few years until the principal is paid off, the payoff could happen in short-order and remain competitive.
Deploying microcell/picocell technology would be easy and could save people like AT&T Mobility/Cingular part of their billions they look to pay for network upgrades. A large scale project here could possibly be done (on-poles) for as low as $44m, and possibly lower as economies of scale come in to play.
I'm hoping someone here reading from GOOG will suggest to any local Ann Arbor Alum (eg: Larry Page) that this would be a chump-change investment that would revolutionize telecommunication in the US.
I scaled my model up to Michigan-size (for fun) and came up with a cost somewhere around 1 Billion to run fiber down every public roadway. Taking the GOOG market cap of ~170Bln, and if I consider Michigan average (don't know, but please stick with me), this could be done for a small part of their market cap, and ROI could be at a reasonable speed. GE and 10GE optics that can do 70km are cheap, sometimes lower cost than that HDTV you just bought, this would make life very interesting...
- Jared
I honestly wonder if they will use ipv4 or ipv6 for their rollout...
Could be interesting to watch!
Maybe they're getting their Ideas from the Irish :). Magnet (www.magnet.ie)
does a similar thing which started over four years ago. They offer fiber to
the home and you can use it for triple-play.
I believe when they started the offering, the bandwidth was (initially
intended to be) limited only by the end user's equipment and they would "pay
as you go" but it appears now as though they have set the limit to 50 Mbps.
There's nothing stopping Google from offering Triple-play with extremely
cheap long-distance calls, Internet, and HDTV. That kind of bandwidth could
easily be utilised, but what next? Google Thin clients? Very exciting!
Regards,
Ken
Jared Mauch wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1Jared Mauch wrote:
I think it's great!
I've been preparing to float a similar idea locally.
If this is how they use their market cap, I would love for them to do it in my local market, which does seem to hold a near-and-dear place in the heart of some google C* types.
- Jared
* Local details/breakdown: http://puck.nether.net/~jared/blog/?p=84
Awesome write up.
Has anyone in the NANOG community been approached by google? I mean
presumably this would require a massive coordination effort with
existing exchange points etc. Or is google going to simply build an
entire long haul network as well? Perhaps combine this with the containers?Thanks. I want to codify it to something more (average) human-readable before I socialize it in the local community.
Sure thing. Make sure to blog it up so folks can contribute feedback 
This sort of investment could have some immediate payback, esp if you have local utility (water, power) buy-in.
Indeed. I was surprised to find how much utility fiber networks exist. I
was in a meet me room in down town Los Angeles, and both So Cal Edison
and DWP had a presence. I knew that DWP had a fiber network, but wasn't
aware SoCal Edison did. Also the city of Burbank power company maintains
a fiber network, which links all the studios together. Unfortunately you
can't bring dark fiber into the major colo there (Qwest IIRC). However
it's quite easy to link any facilities together, and this is heavily
utilized by the studios (most of whom have several sites).
The challenge I see is having the political will to undertake the project.
Hah. Right. Especially with telcos being large campaign contributers.
If you adjust rates up over the first few years until the principal is
paid off, the payoff could happen in short-order and remain competitive.
Mmhmm. And quite frankly, this wouldn't really be necessary if the
telcos actually did last mile build outs of fiber at a decent pace.
People are very willing to pay for this stuff. It's been proven time and
time again. Otherwise the muni folks wouldn't have passed bond measures,
started build out and been sued into oblivion by the telcos. That was
treated as a last resort, after lack of action by the incumbents.
Deploying microcell/picocell technology would be easy and could save people like AT&T Mobility/Cingular part of their billions they look to pay for network upgrades.
Yep. They should become partners in these efforts and help guide the
overall design/requirements etc. Jump in and discuss things like
CoS/QoS/e911 etc etc etc. Not to mention considerable expertise in the
construction of large scale networks. Alas they won't see it that way
A large scale project here could possibly be done (on-poles) for as
low as $44m, and possibly lower as economies of scale come in to play.
Exactly. Especially if the various utility companies can realize the
benefit. Smart grid etc. I have no problem with certain amounts of
bandwidth being reserved for uses by city governments/ utility
corporations who help shoulder the initial build out costs.
I'm hoping someone here reading from GOOG will suggest to any local Ann Arbor Alum (eg: Larry Page) that this would be a chump-change investment that would revolutionize telecommunication in the US.
It sure could. Far more attractive from a CAPex and OPex perspective.
I scaled my model up to Michigan-size (for fun) and came up with a cost somewhere around 1 Billion to run fiber down every public roadway. Taking the GOOG market cap of ~170Bln, and if I consider Michigan average (don't know, but please stick with me), this could be done for a small part of their market cap, and ROI could be at a reasonable speed. GE and 10GE optics that can do 70km are cheap, sometimes lower cost than that HDTV you just bought, this would make life very interesting...
Quite.
- --
Charles N Wyble
Linux Systems Engineer
charles@knownelement.com
http://www.knownelement.com
Hi Jared
You can now nominate your community
http://www.google.com/appserve/fiberrfi/public/options
Regards
Abdul
Wonderful move - might breath life back into the small ISP market. I hope it's a fully multicast-enabled network too.
Antonio Querubin
808-545-5282 x3003
e-mail/xmpp: tony@lava.net
I honestly wonder if they will use ipv4 or ipv6 for their rollout...
Could be interesting to watch!
Hopefully both. This could be one of the first large scale, dual stacked
offerings to end users. There is of course Comcast who recently
announced a v6 beta, and impulse.net for folks in the SoCal region. Not
sure of any other CLEC types offering v6, but if you are speak up!
I guess the phrase innovate/catch up or get run over applies here.
- --
Charles N Wyble
Linux Systems Engineer
charles@knownelement.com
http://www.knownelement.com
announced a v6 beta, and impulse.net for folks in the SoCal region. Not
sure of any other CLEC types offering v6, but if you are speak up!
I suspect you're more likely to find regional ISPs offering v6 than CLECs. The latter seem driven by the sale of circuits and bandwidth, not necessarilly in the efficient or innovative use of those circuits and bandwidth.
I guess the phrase innovate/catch up or get run over applies here.
Yep.
Antonio Querubin
808-545-5282 x3003
e-mail/xmpp: tony@lava.net
Are they going to use Google routers for the deployment?
There are some FTTH deployments in the US, like the well known FIOS
to a number of lesser known municipal deployments in small towns.
If you want to live in a house that is served in this way, how do
you find it. I don't believe there is a "FTTH" field in MLS yet.
Would be nice to have a google maps mashup, or similar...
They don't have a field in the MLS for that, but most people put the
description FTTH in.
There are quite a few communities with FTTH in the Wash DC metropolitan area
that is not FIOS. Openband is one of them serving my house. The 100M fiber
comes into a transition network converter and then to a Netgear. I doubt
that any house would have FTTR (rooms).
I have gig copper ran all over my house. Handy for large file transfers. I have fios as well, and wish it was faster. (yes, all I know is it's a setting, it costs them nothing more)