World’s Fastest Broadband at $20 Per Home | Hacker News Reader

World’s Fastest Broadband at $20 Per Home

bits.blogs.nytimes.com

65 pointshko38 comments

Unfortunately, the article completely ignores the geographic realities or the United States and Japan.

First, Japan's population density is orders of magnitude greater than that of the United States. It's going to cost less to deliver high-speed internet when people live closer together. Likewise, Japan is a smaller country. That 160Mbps is not the speed you're going to get traveling over underseas cables to North America. That's the speed you'll get for servers located within Japan.

Backbone connections are going to cost more to create in the United States simply because Americans are as likely to access a server across the continent as they are within 100 miles of them. Where you can essentially create a backbone in Japan with a under 1,000mi of fiber, in the US you'd need 20x that amount or more.

Those are stark differences. I mean, it wouldn't be too hard for an ISP to provide faster connectivity from Boston to Philly where populations are dense for sites hosted in that area, but that wouldn't be helpful to users - since most of the time I'm not connecting to servers hosted within eastern Mass, RI, CT, the NYC metro area or Jersey. In fact, I'd say that most of the time I'm connecting to Texas or California.

Can we stop comparing Apples and Oranges? The US can never have broadband like Japan and it's not because of corporations or government or any sort of conspiracy. Maybe those things could offer marginal improvements (seeing our service go from 6Mbps to 10-15Mbps), but we simply won't have service like Japan will since we don't locate ourselves close to the servers we want to access or to each other.

The US can never have broadband like Japan and it's not because of corporations or government or any sort of conspiracy.

I accept that geography is a factor, but do you mean to say that lack of competition isn't? It seems to me that it's simply not in the interest of US cable companies to improve bandwidth (which, as the article says, might lead people to stop paying for TV) or lower costs (which would lower their profits).

In a country like Japan, it's decently easy for a company to run its own backbone and provide connectivity to, say, 80% of where a customer is going to be connecting to with minimal investment. In the United States, you don't really have that option. Bandwidth is going to be expensive here. If you up people to 160Mbps service, you're going to need a ton more backend capacity and because you can't just create that capacity yourself, it's prohibitively expensive.

Sure, lack of competition is a problem, but that's just an easy scapegoat. I mean, bandwidth at 10 cents per GB is really cheap bandwidth - cheaper than you'd get with most dedicated hosts. Part of what makes that expensive is that fiber has to go a long way to provide that bandwidth - if I'm downloading from NY to San Fran, that's a huge expanse to cover.

In Japan, it would be a lot easier for companies providing bandwidth to consumers to also run their own backbone.

Sure, competition is a problem in the US, but we're talking about having to do 20x as much cap-ex to cover the US. Anytime you have 20x greater costs, you need to expect that the product will reflect that. Saying that it's just competition is an easy way to be anti-corporate without addressing the issue - it costs a lot more to create and run the internet in the US; it doesn't cost a little bit more or 50% more, but orders of magnitude more because we're more spread out. We don't need to overcome a cost premium of 20%, but probably in excess of 1000% - and it's somewhat incredible that we aren't doing worse.

Dude, the geography and distrubition of population in the US isn't uniform.

Some cities are super-dense. Yet it doesn't mean that they will have really fast internet, since the cost of upgrading is a lot cheaper. What it cost to run line in a community is entirely different than what it cost to run in a city.

Yeah, but generally speaking carriers price things independent of cost. I'm guessing this is because it would be lethal to say, "you live in a less dense area, but we still serve you, just at twice the price."

Plus, if you had read, a large part of the issue is that some places might be dense, but you're not connecting to servers within that dense area. If you live in Manhattan, you might live in a dense area, but the backend cost of connecting you to servers in Texas and California and laying (or paying for) a lot of fiber to do that.

So, there's two costs. The cost of wiring a less dense population and the cost of backend wiring a larger geographic area where servers might be located. Even if you live in NYC, you still have the second cost.

1) The "backend cost" cannot be much more than what it is in Japan. They have to develop those networks as well. You make it sound like all the Japanese are actually accessing US servers. I would guess that the vast majority of the traffic is internal to the country. Yes, the distance is greater, but running "a few" high speed cables between the East and West coast is way cheaper than wiring a whole city.

2) You argue that you can't have one price for densely populated areas and another otherwise. In effect, this is exactly what it's happening, except that in those less dense areas you don't get service at all.

Any way you slice it, you just can't add up the huge difference in what we're getting and what they're getting.

No, his point is that a communications company in Japan can "wire the whole country" and peer with fellow network providers, while building such a comprehensive network in the US is prohibitive, and thus ISPs have to pay for transit.

Basically, a bigger and more distributed (between "autonomous systems") network increases the asymmetry of the interconnections, leading to more billable events (antithesis of "the internet").

> If you up people to 160Mbps service, you're going to need a ton more backend capacity and because you can't just create that capacity yourself, it's prohibitively expensive

Hadn't thought about this factor, and it's interesting food for thought.

But wouldn't that only mean proper ISPs would limit external traffic (unlimited internal p2p) ? Now we're right back to that conflict of interest where the incumbents stall on providing commodity bits because they profit through proprietary services.

This argument is complete BS.

Running fiber lines between large cities is cheap; there's far more bandwidth there than anyone could possibly need at the moment. The main cost is the last mile of connection.

And the last mile of connection is going to cost you pretty much the exact same thing in New York City as in Tokyo. There is absolutely no technical justification of any sort for the United States' backwards internet situation, at least in regards to urban areas.

I'd agree with you for most places in the US vs most places in Japan, but NYC vs Tokyo may or may not be an apt comparison. Putting any lines at all in NYC is a huge hassle. There is basically no clear map of the underground areas. Try getting some quality dedicated lines put in in NYC, it sucks. I can't comment on Tokyo, but it would be a lot easier to run a startup in the suburbs simply for ISP reasons if only the talent would agree (damn fickle trendy programmers).

Japan's population density is orders of magnitude greater than that of the United States.

Japan: 337/km2

United States: 31/km2

http://en.wikipedia.org/wiki/List_of_countries_by_population...

Only one order of magnitude, it appears.

http://en.wikipedia.org/wiki/Order_of_magnitude

And Japan has quite a large land area as countries of the world go. It is larger in land area than most European countries, and is a mountainous archipelago rather than a continental plain. It's hard to say what cost difference between the United States and Japan might be "natural," and what cost difference results from policy differences.

I doubt that everyone in Japan can get Internet that's this fast. I assume that "Japan" in the article means Tokyo or Osaka.

"The industry is worried that by offering 100 Mbps, they are opening Pandora’s box, he said. Everyone will be able to get video on the Internet" - WTF? This really pisses me off!

This is because cable companies are worried that streaming video will put their VOD services out of business. Hence many companies trying to add caps to their bandwidth. When they tell you it's because of 1% of users using the most bandwidth, they're lying; it's because they want users paying for VOD content.

This is understandable, as VOD is by far their most profitable business segment on a %-basis (I work for an IPTV company). Thus, they are scared to death by the concept of decent internet service--it'll reduce their profit margins.

Really? We haven't slipped through a hole in the space-time continuum and been magically transported back to 2003?

There will be decent internet service. Anybody whose business model depends on internet access sucking forever needs to find a new source of income, fast.

Whenever this comes up on HN, my response is that just because it is hopeless having a business model that depends on copyright (crappy internet), doesn't mean they don't benefit from prolonging the process.

The days of mature consumer technologies depending on Broadband speeds you get in S. Korea are a ways away. How many people are still not connected to the internet at all? How many people have no broadband at all?

If they can add 1 or 2 or 5 or 10 years to the point where youtube is what people connect to their big TV, they make a lot of money.

It's sort of like saying a business model that requires U2 to stay popular is not sustainable. Of course not. But it doesn't mean you should give up.

I don't buy this argument. I can easily stream HD video with "only" a 6M DSL connection. (Note that it's re-compressed h264, not mpeg2... but who cares, the quality is great.)

Did I miss it, or did that article not explain how it somehow cost $600 more per American home to upgrade here than in Japan? They touched on competition, but it seemed like that was only for the final service, not necessarily for the supplies. I thought maybe homes were more compacted so it was easier to upgrade multiple homes.

I would assume that the combination of denser residentials and fewer suburbs in Japan has everything to do with the cost difference. When you can upgrade 10 homes in Japan with the same space of 1 home in the US, and there are no sprawling suburbs to lay long runs of cable, it's easy to grasp that per-person price gap.

That's still no excuse for the floundering broadband speeds in this country. I live in Rochester, NY, where Time Warner announced that on top of the "10Mbps" that I never get as it is, they'll be introducing capped plans, with a maximum cap of only 40GB! And I can't just take my money elsewhere, because the city granted them a monopoly; my only other options are horrible DSL that has a 25GB cap, ClearWire broadband at 1/10 the speed I have now, or dial-up. There's absolutely no competition!

It's no wonder Time Warner chose my town for second-round "evaluations"; nobody can leave Time Warner because of this policy, so of course they'll "find" that people are "happy" with the caps. It disgusts me!

> And I can't just take my money elsewhere, because the city granted them a monopoly

Did the city grant the monopoly to help you or because they got paid off?

Someone mentioned torches and pitchforks. Tar and feathers are also useful.

my only other options are horrible DSL that has a 25GB cap

I don't think this is true. If you can get a loop to your house, you can get some sort of business DSL plan, which definitely won't have a cap. I use Speakeasy for this. Expensive, but worth it because I have an unlimited pipe that I can use for whatever I want, and they have actual people (who are native speakers of English) that I can call when something goes wrong.

Anyway, you aren't looking hard enough, or you aren't willing to pay for more than 40G a month.

I live in an apartment. I have no other options...

When you can upgrade 10 homes in Japan with the same space of 1 home in the US, and there are no sprawling suburbs to lay long runs of cable, it's easy to grasp that per-person price gap.

Actually, I've seen a sprawling suburb of Tokyo, and the one resident I visited there was able to get a 100M connection. With the first six months free or something.

The $800-whatever prices quoted in the article are the prices for Verizon's FiOS service. The $20/home is to upgrade Japanese equipment to DOCSIS 3 over existing cable lines. It's different! The article then continues to state that it'd take at the most $100/home to upgrade to DOCSIS 3 in the United States. The section on competition refers to the slothfulness to roll this technology out in the United States at affordable prices.

Why is it that in Silicon Valley, the fastest consumer connection that I can get is 12Mpbs? And that costs me $60 a month.

Sharpen the pitchforks and light the torches.

Comcast recently rolled out (last month) 50 Mbps down / 5 Mbps up, they call it Extreme 50, it's $140/mo.

That's a joke in Japan, South Korea and Escandinavian Nations

Having just spent a week in Iceland, they may have fast local connections, but for actually using US based sites and services, I'd rather have 1.5 MBit DSL in the States (actually I have the Comcast 50/10 service at home). It was painful.

how is the Comcast 50? I am considering getting it, but I'm worried because they advertise it as "up to" 50. What speeds are you getting from it?

I love it! It's $$ but it's a business expense. I get the full 50/10 they advertise: http://www.digitalsanctuary.com/tech-blog/general/comcast-50...

The only issue is you're still limited by the 250 GB/month transfer cap. My first month with 50/10 I did over 750 GB, and got a phone call about it:>

I can't attest for South Korea and Scandinavian countries, but Japan's internet access was unbelievable when I was there for a week on vacation. From what I understand, they're working on getting 1GB of bandwidth to the home over there.

Meanwhile I see more Comcast commercials here telling me how good we have it, but no comparable service to back it up. Torches and pitchforks indeed.

FTTH, right? http://en.wikipedia.org/wiki/Fiber_to_the_premises#Fiber_to_...

I just got 24 Mbps down here in Finland, it costs $66 / month though. I didn't do much price comparison though, I think the same can be had for much less if I switched providers.

I think that you're onto something, namely the, "Why is it that all I can get is X and it costs me Y? Sharpen the pitchforks and light the torches!" meme. So many things this could apply to.

I would gladly pay the $100 to get my home upgraded. Doesn't it seem like individuals or even communities should be able to round up the money to pay for the upgrade themselves? It's basically like we are at the mercy of the cable/telcos until they decide it's cost effective for them to upgrade people. It's not like a hard drive where you can pay extra money to get something better, you have no options.

The $100 is probably an amortized number, assuming that your whole neighborhood gets upgraded. e.g. It costs $50,000 to upgrade your cable node which serves 500 customers. If only 50 people round up the money to pay for the upgrade they'll need $1,000 each, which isn't as palatable.

Amortize that $1000 over two years, though, and the upgrade costs less than $50/mo, which starts to sound reasonable again. If all 500 customers paid, it only takes a price increase of $5/month to pay off the upgrade in two years.

Yeah, I'm a believer in the concept of customer-owned broadband, but the ISPs are stuck in a model where they think they have to accept all the risk of upgrades.

Reply on news.ycombinator.com