Yes, this is why telcos are interested in fiber but cable companies are not. DSL is limited to ~48 Mbps yet in many cities the standard cable plan is 100 Mbps and 300 Mbps plans are available.
Our fiber network has been a huge asset for the local economy.
It helped give birth to our own little tech startup ecosystem.
It made way for the build out of city wide wifi.
It's run by our local power company, EPB, who remains committed to net neutrality, and refuses to enact data caps.
It's forced ATT/Comcast to offer gigabit+ services to remain relevant.
I could go on and on.
Yes, its true that no 'killer gigabit app' has emerged, but give it time. IoT, decentralized services, etc. are all still in their infancy. Its too early to dismiss the internet of the future as a boondogle because 'it doesnt make money' yet.
Seems like you've listed a bunch of things that together might well be considered a killer app. Sure it's not an app you put on your phone and print money with just because it's got access to gigabit. But quantitative differences eventually become qualitative ones.
The main problem with gigabit is that the person on the other end of the line doesn't have gigabit. If someone with 10MBps upload sends me a file, it's still coming in at 10MBps, even if I've got a terabit connection.
Likewise, you can't build an application that requires such tremendous bandwidth unless you know there are enough customers out there who have it. A few small towns here and there aren't going to cut it.
Light up the wires in New York City, and then you'll see some action.
> Likewise, you can't build an application that requires such tremendous bandwidth unless you know there are enough customers out there who have it. A few small towns here and there aren't going to cut it.
From the article
> about 14 percent of Americans have access to gigabit speeds today.
That is quite a bit more than "few small towns here and there", and frankly nearly 50M people (just in US, globally much much more) should be more than enough for building user base.
Those numbers probably vastly inflated (The FCC defines access only by zip-code, and there were many zipcodes with just a few homes given access to broadband to game the numbers for many years), and "Access" is not the same as "having" to many, especially as services that take advantage of gigabit have not yet become available, in a chicken->egg problem.
There's an interesting condundrum - I want my parents to have instant access to a variety of photos and videos that I produce, and for a while it looked like fiber networks were going to be the way - I'd host a home fileserver, and they'd be able to stream whatever they wanted to me from there. But my home internet speeds have actually gone down over the years - Was on a 100mbit connection, and now the plan is '50/50' that rarely achieves more than 10 up.
The cloud has become the distribution network, and it makes a lot of sense in many ways - It also works as a handy backup mechanism, and I trust 'cloud' services to keep good copies of my data more than I trust the second hard-drive I used to keep around.
The killer app is fiber reliability as compared to copper and coax where line errors are the norm, particularly every time it rains and water gets into a compromised circuit.
The speed isn't just a bonus, it means you can do things that weren't possible before, or were painfully slow.
Just as dial-up held us back, but the benefits of broadband weren't realized until we finally perfected things like streaming audio and video, the impact of gigabit connected networks will take time and may come in an unexpected form.
The more immediate effects will be that distributed computing becomes no big deal, and accessing your files from the cloud or from a home device while remote becomes frictionless. Instead of necessarily lugging around a laptop you might travel lighter, confident that you can get access to what you need anyway.
It also makes apps like Dropcam possible where you stream endless hours of video to a remote server on the off chance you might need it. The cost of maintaining multiple streams becomes so low you don't even worry about it. No longer do you need to fret over the equivalent of popping a breaker when trying to watch Netflix as well.
The funny thing about these applications is they don't seem like a big deal when you have them, but when you suddenly lose them it's a huge problem. This is much the same way we take electricity and running water for granted, never thinking much of it, but when it cuts out we're in trouble.
I've had gigabit internet (symmetrical) in my apartment in London for the past ~3years.
Some things I've found:
Even 802.11ac WiFi is really lacking. Despite link speed being 900mbit, it really struggles to push more than 200mbit/sec. On 802.11n it was more like 90mbit/sec.
Torrents are often IO (or even CPU) bound at high speed. I use a transcend drive on my MBP for more space and that seems to max out at 100mbit/sec (yes, bits) with very random read/write like bittorrent.
You end up not noticing internet faster than 100/100. I have 0.3ms pings to the local Cloudflare CDN, so latency is not a problem. You do notice slower than that. Especially low upload connections, you are so used to git push taking no time at all, you forget on slow upstream how slow that can feel.
While it's nice to have gigabit internet, I would rather 100/100 with low latency (eg: GFast) over 300/30 with variable latency (eg: DOCSIS3).
The highest non-4K (I don't have a 4K TV) bitrate I have seen is netflix at around 8mbit/sec, which looks great.
I honestly do think BT in the UK is making the right option in choosing a faster/larger deployment of FTTC and GFast over FTTH. Gigabit really isn't that useful, and I can't see this changing for the next few years.
Fiber in the ground is a long term investment. The old copper might have been there for decades, and it is getting pretty much a dead end as far as tech goes. Sure, gigabit now seems overkill for average consumer, but how about in ten years time? How about twenty years?
Of course those decade old copper networks do not just magically maintain themselves in perfect order. I believe that in many fiber deployment cases the local network was already due for some sort of major operation, due just general flakiness, low speed, or poor availability. Fixing that is expensive no matter how you do it, and it really wouldn't make sense to lay down new copper anymore these days if you can avoid it.
There is one major consumer application for gigabit-scale internet: self-hosting. Gigabit internet is one key factor in enabling people to move away from massive clouds to personal services. Combine this with the common NAS boxes, device whose popularity is already being fueled by light ssd-powered laptops, and it is not that farfetched to see something like sandstorm to really take off.
I think the point of the article is questioning when to make that investment. If you're Verizon and Sandy destroyed your copper, sure, go ahead and replace it with fiber. If you're a cable company with a network that currently works fine, fiber mania is a waste of money today. There's some benefit in delaying since the routers needed to light that fiber will be somewhat cheaper in the future.
Also, Sandstorm runs much better in a datacenter than in your home.
The problem with people not knowing how to utilize gigabit networks is that it's sort of a "cart before the horse" type of problem.
Until enough people actually HAVE symmetrical (both up and down) gigabit speeds, you will be limited by the transmission speed of whichever provider you are downloading from.
A lot of web servers are sitting in racks in data centers, and you will only ever be able to download a file from that web server at the speed that that web server is able to send. Assuming the web server has a single gigabit connection, and ten other people are downloading from that server at the same time, you might only be getting 50-100 megabits tops.
Where gigabit networks for consumers would REALLY start to shine is when thousands of people are on gigabit, and are all doing peer-to-peer transactions, or hosting their own servers.
Imagine if there were 100,000 people with decent computers (at least a terabyte of drive space apiece), with every computer connected via fully bi-directional gigabit speed connections.
Now imagine if you were to build a fully encrypted peer to peer file storage cloud network on top of that, with each client having their own key (which only they had access to). Encrypted data would be stored in pieces across the cloud, with redundancy and parity correction stored across nodes.
If you built something like that properly, you would have a giant, disaster resistant cloud drive, where you could store tons of private data and access it from anywhere at gigabit speeds. If you had a large file that was striped in redundancy across a thousand nodes, you could even achieve terabit speeds by downloading the pieces in parallel and assembling them on your end.
You could even tie a pricing model into it (perhaps powered by a blockchain methodology), where you had to pay monthly to store your data in the peer to peer cloud, and people who contributed storage and machine resources would receive payments in return for their contribution.
You could even take it a step FURTHER and add a cloud processing layer on top of it all, where people could dedicate a couple of their CPU cores for usage by the cloud (when their machine wasn't in use), and they would be compensated.
If you had that, you would have the ability to not only store data in the cloud, but also process the data (and even do things like run distributed web servers).
The possibilities are endless!
But it will never happen until it makes sense, and for it to make sense, there needs to be a critical mass of people who actually HAVE symmetrical gigabit connections.
Gigabit is backward compatible to any slower speed. The unused bandwidth is not wasted somehow. The current copper network is going to have to be replaced anyway. It doesn't make a lot of sense to run fibre past buildings just to connect more and more VDSL2 and DOCSIS 3.1 boxes sitting on the boulevard in a desperate attempt to squeeze another decade of life out of shorter and shorter lengths of existing copper.
I think the killer app for gigabit symmetrical internet service will be the ability to host your own apps (from social networking to video streaming) either on the router or another appliance. It's just hard to see how that will ever become a common practice until someone builds such appliances that are friendly with each other beyond their primary manufacturer because the natural inclination of any corporation making such appliances is to make them locked in. So it'll be interesting to see if that pattern keeps up or softens.
20 comments
[ 3.7 ms ] story [ 49.0 ms ] threadOur fiber network has been a huge asset for the local economy.
It helped give birth to our own little tech startup ecosystem.
It made way for the build out of city wide wifi.
It's run by our local power company, EPB, who remains committed to net neutrality, and refuses to enact data caps.
It's forced ATT/Comcast to offer gigabit+ services to remain relevant.
I could go on and on.
Yes, its true that no 'killer gigabit app' has emerged, but give it time. IoT, decentralized services, etc. are all still in their infancy. Its too early to dismiss the internet of the future as a boondogle because 'it doesnt make money' yet.
Likewise, you can't build an application that requires such tremendous bandwidth unless you know there are enough customers out there who have it. A few small towns here and there aren't going to cut it.
Light up the wires in New York City, and then you'll see some action.
From the article
> about 14 percent of Americans have access to gigabit speeds today.
That is quite a bit more than "few small towns here and there", and frankly nearly 50M people (just in US, globally much much more) should be more than enough for building user base.
There's an interesting condundrum - I want my parents to have instant access to a variety of photos and videos that I produce, and for a while it looked like fiber networks were going to be the way - I'd host a home fileserver, and they'd be able to stream whatever they wanted to me from there. But my home internet speeds have actually gone down over the years - Was on a 100mbit connection, and now the plan is '50/50' that rarely achieves more than 10 up.
The cloud has become the distribution network, and it makes a lot of sense in many ways - It also works as a handy backup mechanism, and I trust 'cloud' services to keep good copies of my data more than I trust the second hard-drive I used to keep around.
The speed is just a bonus.
Just as dial-up held us back, but the benefits of broadband weren't realized until we finally perfected things like streaming audio and video, the impact of gigabit connected networks will take time and may come in an unexpected form.
The more immediate effects will be that distributed computing becomes no big deal, and accessing your files from the cloud or from a home device while remote becomes frictionless. Instead of necessarily lugging around a laptop you might travel lighter, confident that you can get access to what you need anyway.
It also makes apps like Dropcam possible where you stream endless hours of video to a remote server on the off chance you might need it. The cost of maintaining multiple streams becomes so low you don't even worry about it. No longer do you need to fret over the equivalent of popping a breaker when trying to watch Netflix as well.
The funny thing about these applications is they don't seem like a big deal when you have them, but when you suddenly lose them it's a huge problem. This is much the same way we take electricity and running water for granted, never thinking much of it, but when it cuts out we're in trouble.
Some things I've found:
Even 802.11ac WiFi is really lacking. Despite link speed being 900mbit, it really struggles to push more than 200mbit/sec. On 802.11n it was more like 90mbit/sec.
Torrents are often IO (or even CPU) bound at high speed. I use a transcend drive on my MBP for more space and that seems to max out at 100mbit/sec (yes, bits) with very random read/write like bittorrent.
You end up not noticing internet faster than 100/100. I have 0.3ms pings to the local Cloudflare CDN, so latency is not a problem. You do notice slower than that. Especially low upload connections, you are so used to git push taking no time at all, you forget on slow upstream how slow that can feel.
While it's nice to have gigabit internet, I would rather 100/100 with low latency (eg: GFast) over 300/30 with variable latency (eg: DOCSIS3).
The highest non-4K (I don't have a 4K TV) bitrate I have seen is netflix at around 8mbit/sec, which looks great.
I honestly do think BT in the UK is making the right option in choosing a faster/larger deployment of FTTC and GFast over FTTH. Gigabit really isn't that useful, and I can't see this changing for the next few years.
Of course those decade old copper networks do not just magically maintain themselves in perfect order. I believe that in many fiber deployment cases the local network was already due for some sort of major operation, due just general flakiness, low speed, or poor availability. Fixing that is expensive no matter how you do it, and it really wouldn't make sense to lay down new copper anymore these days if you can avoid it.
There is one major consumer application for gigabit-scale internet: self-hosting. Gigabit internet is one key factor in enabling people to move away from massive clouds to personal services. Combine this with the common NAS boxes, device whose popularity is already being fueled by light ssd-powered laptops, and it is not that farfetched to see something like sandstorm to really take off.
Also, Sandstorm runs much better in a datacenter than in your home.
Until enough people actually HAVE symmetrical (both up and down) gigabit speeds, you will be limited by the transmission speed of whichever provider you are downloading from.
A lot of web servers are sitting in racks in data centers, and you will only ever be able to download a file from that web server at the speed that that web server is able to send. Assuming the web server has a single gigabit connection, and ten other people are downloading from that server at the same time, you might only be getting 50-100 megabits tops.
Where gigabit networks for consumers would REALLY start to shine is when thousands of people are on gigabit, and are all doing peer-to-peer transactions, or hosting their own servers.
Imagine if there were 100,000 people with decent computers (at least a terabyte of drive space apiece), with every computer connected via fully bi-directional gigabit speed connections.
Now imagine if you were to build a fully encrypted peer to peer file storage cloud network on top of that, with each client having their own key (which only they had access to). Encrypted data would be stored in pieces across the cloud, with redundancy and parity correction stored across nodes.
If you built something like that properly, you would have a giant, disaster resistant cloud drive, where you could store tons of private data and access it from anywhere at gigabit speeds. If you had a large file that was striped in redundancy across a thousand nodes, you could even achieve terabit speeds by downloading the pieces in parallel and assembling them on your end.
You could even tie a pricing model into it (perhaps powered by a blockchain methodology), where you had to pay monthly to store your data in the peer to peer cloud, and people who contributed storage and machine resources would receive payments in return for their contribution.
You could even take it a step FURTHER and add a cloud processing layer on top of it all, where people could dedicate a couple of their CPU cores for usage by the cloud (when their machine wasn't in use), and they would be compensated.
If you had that, you would have the ability to not only store data in the cloud, but also process the data (and even do things like run distributed web servers).
The possibilities are endless!
But it will never happen until it makes sense, and for it to make sense, there needs to be a critical mass of people who actually HAVE symmetrical gigabit connections.