@craig: i am disappointed. you stoop too low to collect scandalous brownie points. this is bad reporting given there is absolutely no basis for this assertion, except for a rambling blogger who has no proof.
on the other hand, a very credible piece of information was released by a noted analyst that ciena has just won the at&t optical domain, and i see no mention of this potential path-breaking news.
@eve: holy smokes! only logic i see here is: google = cheap, cheap = huawei, hence huawei wins?
odo <- who loves sensationalism, but expects an iota of credibility from lr
@balags77:If you mean by that it was an April Fool's day, sorry but no. Just google around a little and you'll see. I know first hand that these servers are used in the datacenters.
What kind of right-of-way permit do I need to dig up city streets and put fiber in? I'm thinking of running a community FTTH service so I can offer high-speed access to my home videos to everyone in my neighborhood.
Jepbjr has a good point -- that the planning and city-engineering kinds of concerns tend to be huge in building infrastructure.
Teresa Mastrangelo, analyst with Broadband Trends, was telling me Google could choose to get around that. It's going to hand-pick its munis, so why not pick the ones that give an auto-green-light to the project and/or are willing to even pay for some of it? The latter part might be tricky given the recession, but I like her point: Google can take a shortcut. They don't have to go through everything a normal telco goes through.
(my previous post mentioned the possibility of Google finding ways to innovate on deployment costs... well, i suppose that's one way...)
First if all, Google are talking about building point-to-point (P2P) fiber, not PON, right?
Secondly, the Google way, it seems to me, would be not to buy from any of the incumbent vendors at all. Rather, they would buy dirt-cheap consumer-grade mini switches and stack them mile high.
If they would opt for actually buying working gear, Packetfront has been building networks around exactly this concept for years. In Europe, there are hundreds of municipalities doing this since the beginning of the last decade. It works quite well technically, but seldom commercially.
But I think Google will build their own stuff. Is it a good idea? Most probably not. Any kid can build a server, frankly, but building telco equipment according to all the rules and legislation is another beast entirely. Much of the stuff in a telco switch may seem unnecessary, but even the necessary stuff can keep a seasoned development team busy for years. If I were running Google, I'd buy a company like Packetfront instead. The bill would be tiny to what all the construction companies will charge anyway. I dont need to remind any readers of this site that it 's not the switch costs that have kept FTTH development slow.
People tend to think that just because Google do a lot of great stuff, all they do is great - a halo effect. But 98% or so of their revenue still comes from the search engine they launched in the previous century, and all their best stuff since then has mainly imrpoved the search engine business. So, the question everyone should ask themselves is - how is google intending to support its advertising services with this new FTTH service?
I agree that building a metro switch is a different story, but look at Free in france with their home-built DSLAMs. If Google would just put their switches in their own enclosure and just provide residential services, there are really not too many standards to comply to.
Operators usually demand the standards as some sort of warranty.
In any case, I agree that going "homegrown" for this kind of deal probably means too much overhead.
@craig: i am disappointed. you stoop too low to collect scandalous brownie points. this is bad reporting given there is absolutely no basis for this assertion, except for a rambling blogger who has no proof.
on the other hand, a very credible piece of information was released by a noted analyst that ciena has just won the at&t optical domain, and i see no mention of this potential path-breaking news.
@eve: holy smokes! only logic i see here is: google = cheap, cheap = huawei, hence huawei wins?
odo <- who loves sensationalism, but expects an iota of credibility from lr