I want to criticize Google Fiber, but I will give them this: they proved gigabit internet was completely possible to offer to consumers directly to the home, inexpensively. They proved there was demand for it, that there was a potential business here.
Yesterday, I had a 1.5Gbps fiber internet installed in my home- not by Google, of course, but by someone who copied their model.
I think that was the point all along. Move everyone forward on this. Force every other ISP to stop resting on their laurels and get to work pushing the boundaries.
Unrelated: the bottleneck on my bandwidth is, for the first time in my life, between my computer and the modem rather than between my modem and the internet. It's 30,000 times faster than my 56k dial-up was in the 90s. This is pretty neat.
Did they prove that? Even demand did they actually prove customers will pay *more* for fiber?
What it proved is that if cable companies offer lower prices than google, and offer 'good enough' speeds. That people will not switch over. Fiber is fast but so is DOSICS4.0. What it showed was that fiber to the house was not a revenue or profit winning strategy, with instead the ISPs using the existing copper coupled with fiber close to the home.
ATT and others recently made a push for fiber to the home, but their DSL was falling behind coax/cable and they were forced to make the switch.
Personally, look at T-Mobile home internet & starlink's growth. We might see cellular internet to the home take over in the next 10 years. My 5G cellular speed is already nearing that of what an 802.11ac wifi network can deliver. Most people are currently fine with 50Mibps, which is well within 5G's capabilities.
I live in San Antonio, a city where Google Fiber has been an option to a good part of the city for many years. When they came to town standard minimum speeds of competitors bumped from 20-50mb/s to 100-300mb/s almost overnight. Meanwhile, my family and friends just a couple of hours away in Houston were stuck with baseline options in that original range for years until AT&T rolled out their own 1gig offering much more recently.
If nothing else they have proved the competition was needed.
In my neighborhood, I was restricted to 100Mbps down for 10 years @ something like $100->150/month. The price was also constantly climbing. The only competitor for my neighborhood offered 40mbps down for $100/month.
The MOMENT a new fiber provider announced they were going to build out to my area, all the sudden 600 down cost them $60/month and the caps went away.
Once the buildout happens in my neighborhood I'm switching anyways out of pure spite.
Google Fiber completely shit the bed in Louisville (KY). The initial announcements did seem to spur more build-out and price cuts among the two incumbents, but even with One Touch Make Ready, Google eschewed the utility poles they were supposed to use, attempted "micro-trenching" which made zero sense in the city's climate and screwed everything up so badly (including creating a massive mess of potholes for the city to clean up) that they just shut the network back down and skipped town.
Unsurprisingly the incumbents are back to their usual complacency and their prices have steadily been on the rise in the couple years since Google Fiber did such a terrible job. It's one thing to shut down Messenger App 2 of 6, it's another thing to just entirely cut and run from an infrastructure project leaving bills behind to the City's taxpayers to pay.
5G doesn't scale: no mobile tech scales. It works well when it's new, and in remote areas where it is not congested, but whatever speed you can pull out of it doesn't work in cities when people use it as a primary internet connection.
The laws of physics have not changed and yet every new mobile iteration cycle someone wants to insist they have.
Starlink is more akin to fixed wifi, which is essentially a point to point link depending on line of sight. That's the one possible game changer here: it's low enough, and the other end is somewhere you're unlikely to be cut off from.
But I suspect even Starlink, at any reasonable satellite density, wouldn't be able to support a city's worth of internet connections - and any reasonable effort to do so would be plain better served by just installing terrestial cables.
I remember 12 years ago, when my phone first got 4G, it was blazing fast. I was getting 50Mbps download speeds. Eventually it slowed down as more people got on the network. Fast forward to today and my phone has 5G. It’s blazing fast. I’m getting 50Mbps download speeds!
> The laws of physics have not changed and yet every new mobile iteration cycle someone wants to insist they have.
MIMO gets past Nyquist and directional antennas can be added in finer granularity as more users come online (they do this for home LTE, not offering until adding more directional transmitters).
Sure: but the point is that whatever tricks you can pull, you're just slightly bumping the capacity of what is fundamentally a shared channel.
We already have ubiquitous wireless internet: it's wifi. And it gets faster every year, provided we deploy a base station every 10m with fiver backhaul (aka your regular home internet connection).
Directional can scale enormously vs omni directional. And MIMO can scale to a ton of antennas, especially in something like home internet where device size isn't a limit.
Do you think it's practical to put dozens of antennas on every residential property?
That's the point: with enough effort you can make any sort of wireless link pretty arbitrarily large and more "wire-like". At an extremely disproportionate increase in cost of equipment and sensitivity of antennas to positioning.
But now you're competing with the installation costs of fixed cabling anyway - we're not talking about small portable devices any more which a consumer puts "wherever" in their house. We're talking large fixed installations which need to be aimed and calibrated correctly, and consume substantial power (the Starlink rectenna for example is not a low power device).
The moment you've got skilled labor having to come out to the property you've lost: it's just as easy for the skilled labor to install fixed line cabling, as well as less disruptive to the home owner and it is very likely a lot cheaper. Our cities are well setup to bring things into people's homes - we have water, gas, electricity lines that need to go everywhere. The cost is not in the device, it's in the labor.
> Do you think it's practical to put dozens of antennas on every residential property?
Yes, high end cellphones already have 8x8 mu-MIMO. It can be in an array and isn't like putting 12 separate antennas in different places on your roof or anything. Starlink (phased array, different tech) has hundreds of antennas (https://www.youtube.com/watch?v=iOmdQnIlnRo&t=36m30s).
You might be thinking of like TV antennas where they often were much bigger due to the longer wavelengths involved in VHF.
Even if it were big, many many rural homes used to have those crazy big 80s/early 90s c-band satellite dishes just to get more TV channels, and for someone with a poor internet connection I could see them even putting something like that big up if it were required.
There's no real physical limit right now because we have a LOT of mmWave frequency up in the 26, 28, 39, and 41GHz bands. The problem is that the higher the frequency the less propagation you get, the more often you need a base station, the more the network will cost.
The problem on the low band is that you can eventually you get to a point where the cell is overloaded and you're at capacity but still you can throw in another cell and/or split the sectors up. Which, again, costs money.
The problem is essentially economic, not technical.
What happened was that everywhere Google Fiber went, the big ISPs would suddenly offer competitive pricing and service (which of course they could have done all along), which meant it was hard to get customers.
Also, they would fight tooth and nail to prevent Google from being able to actually build the infrastructure. One way was to make it needlessly onerous, slow, and expensive to access utility poles.
This is all straight from the Standard Oil playbook.
>Also, they would fight tooth and nail to prevent Google from being able to actually build the infrastructure. One way was to make it needlessly onerous, slow, and expensive to access utility poles.
It's funny how Americas look down on "third world" countries as being hopelessly corrupt, when in reality America is just as corrupt as any of them. In highly developed nations with rule of law, they don't have these problems.
I have Google Fiber, and I can honestly say that gigabit speeds are largely useless, most websites you connect to and download from are going to bottleneck you at way lower speeds with their upload throttling. It's exciting when I actually get to download things at unrestricted speeds.
So why do I stick with a $70/mo service that I can barely utilize over "good enough" speeds that I wouldn't notice the difference of for $10-20 less per month?
Reliability, and a good business model. I had my internet go out overnight when I wasn't using it and didn't even know I had an outage for 5 hours. My bill is how I found out I lost service. Google Fiber will automatically stop charging you and discount your bill if an outage is caused by their end. No other ISP does that that I know of. You're paying for "speeds up to XXmbps", including speeds less than, and no service at all. I have never once had to call or contact their support. I don't even know what number to call, or off the top of my head where to find that number, and I don't have to, I can actually drive to a Google Fiber physical store and talk to someone in person, not some level 1 tech support agent in India reading verbatim off a script in an internal wiki. Granted, I can also call or email them, but I've never once in over six years of service, had a need to do that. Another thing is, you can cancel and reactivate your service on the go. Going on vacation for a few weeks? Disable your service before you go, turn it on when you get back home, save money, and they don't care. You can shut it off every night before bed and turn it back on when you get home. They allow you to do this. You'll only pay for the service you are actively being provided.
They created a business model that is so polar opposite of what other ISPs are, to push other ISPs to fix their craptastic infrastructure or get left in the dust by competition.
I personally will always choose a fiber connection over cable because I don't know of a single Cable ISP that people don't regard as "a horrible company with equally horrible service and customer service".
If all your neighbors get on 5G with you, you are no longer going to get anywhere near the speeds you currently have. The real benefit of 5G for the phone networks is the higher frequencies means the signals don't penetrate as well, cover a lot smaller of an area which means they can have more radios - to provide more capacity without overlapping.
It also means 5G only makes economical sense in the densest of areas. 5G isn't coming to the 'burbs any time soon. When you get into the one radio per couple of houses you might as well just run fiber.
Universal wireless internet is a fantasy - there simply isn't enough capacity over the air - just look at the growing pains starlink is already experiencing in areas of just moderate density. Wireless has it's place, but it will never be sufficient for base load.
> The real benefit of 5G for the phone networks is the higher frequencies means the signals don't penetrate as well, cover a lot smaller of an area which means they can have more radios - to provide more capacity without overlapping.
Nope, density is manageable in most frequency ranges. The advantage to the mmwave frequency bands is that they have larger channel sizes (up to 400MHz), which allows for faster speeds.
It's cheaper to place a single 5G cell antenna for every 10 homes than to run FTTH to 10 homes. Home users don't need 24x7 bandwidth and generally have sporadic bursts making it possible to oversubscribe.
> Did they prove that? Even demand did they actually prove customers will pay more for fiber?
That's a good point, tbh. We have a local provider that provides fiber to the home. Its high quality and the bandwidth is reasonable for the price. Its also fully symmetric and everything is out of sight (buried underground).
A national cable company came into the neighborhood offering copper and ran new copper with ugly boxes everywhere for a slightly lower price. Quite a few people switched.
> did they actually prove customers will pay more for fiber?
Who said fiber will inherently cost more? That assumption only makes sense if you think that ISPs charge people at a reasonable consistent margin, and arent't just jacking up the prices up to the maximum they can get away with.
From my personal experience, finer didn't end up costing more in most cases. I am paying about $60/mo for a symmetric 1Gbps fiber from Google Fiber in a high cost of living city in the US. No data caps, no random fees, easy billing, easy to work with customer service. I paused the service once because i was out of state for about a month, and then resumed it again. It was as easy as calling their customer service and spending less than a minute on it (not counting the little time at the start of the call where I had to verify my account information). Transferring service to a new address (within the same apt building) was extremely easy. No lock-ins, month by month, so I can easily cancel any time and then sign up later. No "installation" fees or any other bs. Just paying the agreed upon monthly cost, plain and simple.
Just before that, i used to live in another large city with a cheaper cost of living (Atlanta). I was paying about $20-30 more per month for a locked contract with Xfinity. It was a nightmare. Data caps, them adding "by mistake" equipment rental fee to my bill about half a year into it (despite me using my own modem + router, and me giving them their equipment back on day 1), as well as me wasting hours to get it resolved, only for the fee to get added back to the bill a few months later. And I got lucky that me and my roommate at the time spotted that fee getting added sneakily both times. And don't get me started spending over 4-5 hours on the phone trying to get the service canceled a couple years after living there (because i was moving across the country). I legitimately ended up having to lie to them that I am moving overseas, otherwise they kept insisting they could just transfer the service to my new address, no matter how many times I told them that I wasn't interesting.
An interesting development came shortly after. After I moved out of Atlanta, Google Fiber finally arrived into the city. In the areas of metro ATL where Google Fiber entered, Xfinity dropped prices significantly, but they barely matched Google in cost (but still had tons of random bs fees and costs that made the price higher), still was a locked contract, and the speeds/service was much worse, it was closer to 100Mbps (it wasn't fiber, it was DOCSIS). Guess what happened in the metro ATL areas that were just right outside of Google Fiber coverage? Instead of dropping the price, Xfinity jacked up prices there for the same old crappy service.
As soon as I initially heard that Google Fiber was coming into ATL area, I was excited for my mother finally getting a cheaper and more reliable internet provider that wouldn't sneakily add random fees or "deals" that force her to deal with it. At worst, I expected that Google Fiber won't expand to her house (just a 30 min drive up north from ATL midtown), but with Xfinity at least giving a cheaper rate, given the competition was encroaching. Except the exact opposite ended up happening.
Oh, and to make things better, that apartment building I lived in back then didn't get Google Fiber for years since it became available in the city, despite being squarely in the coverage area. Guess why? The apartment building management made a 5-year exclusivity contract with Xfinity (aka Xfinity paid them off) to be the sole ISP offered to the apartment building. So even if I wanted to switch, and Google Fiber was available in my area, then too bad.
Because they stopped. Because they made it sound like everyone was getting Gigabit internet by 2018 and then just... didn't do anything. They had first movers advantage and were offering something incredible.
And they gave up.
It's nice that they're getting back at it now, but come on, where have you been all these years?
In Louisville (KY) Google Fiber was handed every advantage they needed on a silver platter (One Touch Make Ready) versus the incumbents. Decided to "experiment" and did the absolute dumbest thing possible ("micro-trenching" in a state with an active climate and this thing called "Winter"), failed incredibly hard, and just shut down the network, cut and ran rather than live up to any promises they made to the city. (Leaving the bills for things like giant potholes to the City and its taxpayers.)
That wasn't incumbents fighting them. They did that to themselves.
Google couldn’t wire up their own back yard! You think Comcast has more lobbying clout in Silicon Valley than Google? Google rejected proposals from places like Baltimore and LA that were trying to kick out Comcast and asking Google to wire them up.
I was always surprised that even living in Silicon Valley I was never able to get fiber! I’m near San Francisco now and it still isn’t an option to my home.
Because it's an example of Google's pathological ADHD. I'd be hesitant to subscribe to Google Fiber because it's even odds as to whether they'll just decide to shut it down.
That makes no sense. If they shut down Fiber, the worst that’ll happen is you're forced to switch to a worse provider that costs more. Why not get the good speeds and the low cost while you can?
My Modem can support 1.5 gbps to the internet. The cat5e cable between the modem and my router can support 1.0 gbps (rated to that at least), and the one between the router and my desktop is the same.
Over wifi, I can get around 600mbit to the router at best.
Running generic speedtest websites, I'm seeing myself top out at around 900mbit or so. Which should be enough for all my needs for long time.
Yesterday, I had a 1.5Gbps fiber internet installed in my home- not by Google, of course, but by someone who copied their model.
I think that was the point all along. Move everyone forward on this. Force every other ISP to stop resting on their laurels and get to work pushing the boundaries.
Unrelated: the bottleneck on my bandwidth is, for the first time in my life, between my computer and the modem rather than between my modem and the internet. It's 30,000 times faster than my 56k dial-up was in the 90s. This is pretty neat.