When we got fiber, our internet didn’t speed up. We just saw a monthly rate increase. The providers here throttle it so significantly that there isn’t any benefit, not even for upload. The plans that cost an extra $150 per month have far higher limits but that’s insane
I'm on Telus gigabit fibre, for $69/mo. I've never once had issues with throttling, even when I was on Shaw (the 300mbps plan). What kind of throttling are you getting and in what use cases?
I'm on telus fibre, it's only fast within telus' local network because their peering is SHIT. So unless all you're going to do is speedtest all day long telus isn't going to help you.
Canada us always a safe bet for extraordinarily bad internet speed, connection reliability and price. In my experience, our ISPs would make Comcast look like saints. (I'm looking at you, Bell)
I'm far from an expert on L3-L2 ISPs and their tech, but they usually rent the line from whoever owns it and then sell it back to you - then they cap the speed between you and the box.
It's most likely an operational mistake - have you called and asked for them to nuke the line and then set it up again?
Or are you saying that you are getting your advertised speeds, and those just didn't go up when you "got fiber"?
Don't know which one is more worrying if true, to be honest.
I'm sure there are a lot of other bottlenecks as well. I have a pretty run of the mill mid-tier cable plan for $100/month in the US and, assuming it's working like it's supposed to, I've pretty much never been on another connection anywhere including in company offices that made me go "This is so much faster than at home."
Not convinced by the use case section. Very few servers will allow you a full 25gbit download, let alone anything more than a 1gbit (and often less). And if you own the server on the other end, that sort of bandwidth comes at a cost.
I think beyond 1gbit, the benefit become super marginal and the hardware expensive.
why stream on demand if we have the bandwidth to just send the whole video in the first 100ms? this might actually save power; instead of back & forth back & forth with the service, we can transmitnthe whole thing & be done, the server can now go serve other people.
I can't find it right now but I remember reading a paper by Cerf? (I'm not great with names) that detailed such a concept. If I remember it I will edit and post a link.
Multiple 4K streams become possible. E.g. when the whole family comes over for the holiday we can all ignore eachother in favor of 4K video streams (which have better resolution, color, and cinematography than the real world ;) )
There aren't many (any?) services that are offering Bluray quality 4K HDR streams but even those are only 140 mbps. I would think 1 gigabit down should cover most families handily, especially when you consider most streaming services are more like 20 mbps max.
The main argument I see is "if you build it they will come" i.e. we won't see higher bitrates until connections like these are more common, but for now gigabit has even fairly extreme use cases for video streaming pretty well covered.
Even where the ports are available, 25 Gbit from a single address is well into the realms of looking like attack traffic in a wide variety of scenarios.
Past even 500 Mbit I'm way more interested in latency considerations than raw bandwidth, and practical matters like how to use that bandwidth from my laptop (good luck doing 500 mbit wireless reliably, never mind 25 Gbit!)
I have a bandwidth in the low hundreds and it definitely feels excessive to me. But changing from wireless to wired and getting rid of those occasional latency spikes - very noticeable while playing games. A static ip is also something I would want.
Most routers and wifi adapters are crap. Buyers do rarely go beyond "wifi 5" or "wifi 6", and do not realize that there's much more.
The older Apple Macbook Pros (pre-2019) came with 3x3 MIMO ac adapters. If you had capable AP on the other side, you could reliably do gigabit with them. The newer ones have only 2x2 MIMO, just like the rest of the laptop market, so you will get only 600-700 Mbps (out of theoretical 866 Mbps).
If you are getting 500 Mbps and there's not a concrete wall between your client and the AP, something is quite wrong. Misconfigured AP, your client cannot do multiple streams (yes, there were adapters like that sold on the market), or just older/pre-ac AP or client.
Basically, how fast you can transfer over wireless is determined by:
1) how effectively you can pack data into a channel, determined by the modulation scheme.
In your wifi properties, you can see the modulation scheme used as 'MCS index'. You will see an integer between 1 and 11, the higher, the better. This is negotiated between client and AP, depending on the signal strength (antennas, number of walls/wall material between them, etc). With wifi 5, it is realistic to have MCS 8 or 9, if your client and AP are in the same room.
2) how wide is your channel - the "basic" is 20 MHz wide, newer standards can combine multiples, into 40, 80, 160 MHz wide one. Wifi 7 will bring 320 MHz wide channels.
2,4 GHz band has only 3 non-overlapping 20 MHz wide channels available. If you use more than one, your neighbors will hate you. The situation is better with 5 GHz band, unless you run into the DFS (radar detection), then you can have intermittent outages. This is improved by Wifi 6e which brings more channels in the 6 GHz band, but the client and AP support is not there yet. Currently, running an 80 MHz wide channel bellow DFS frequencies is probably the best.
3) how many many simultaneous streams you are able to transmit/receive - MIMO (one client using multiple streams), MU-MIMO (multiple users using multiple streams, so they do not stomp on each other and sidestep next point).
Most clients are capable of 2x2 MIMO, older Apple MBPs (pre-2019, -ac based) are capable of 3x3. Most APs are also 2x2. The nicer APs can do 4x4 MIMO, some gamer APs by companies like ASUS can do 8x8.
4) how effectively can multiple clients share the same channel.
Here the problem is, that once an older client connects, it is downgrade for all clients, they must be compatible. If you can, keep your 2,4 GHz network -n and higher, and 5 GHz -ac or higher.
Once you will be able to have a band -ax only, you could theoretically be able to use OFDMA, which does improve the performance for multiple clients.
You can have a look at mcsindex.com: pick the modulation scheme (row), number of MIMO streams (group of rows), how wide is your channel (column) and OFDM/OFDMA (group of columns) and there you can see the theoretical bandwidth. Note, that it is a shared bandwidth, for all clients.
There are some minor points:
5) if you have IPTV or another application that uses multicast, do not run it over wifi! Multicast by definition does not have ACK, so what can AP do to ensure every subscribed client gets the data? Slow everything down, for everyone. So do not let it do it.
6) multiple SSID - possible, but do not overdo it; they also have performance impact (the SSID advertising takes time, at the slowest/most compatible rate allowed; the more of them, the more time it takes). Ubiqiti limits them to 4 per radio.
7) at higher frequencies, you won't have a such range than at lower ones. This is both good and bad thing. The bad, obviously, that your range is lower. The good is, that the range of your neighbors is also lower, so you could use the same channel without disturbing each other.
8) It does not help to crank the tx power to the max. The clients may receive, but when they answer, the answer could never arrive. Clients do not have antennas or tx power like APs do. Use the tx power reasonably; if you have multiple APs in your house, adjust it so clients will let go the weaker signal and connect to the stronger one (which AP connect to is fully managed by clients, unfortunately, and they often prefer to stick to the one they are already connected, even if different one with stronger signal is available. At most, some APs can kick the client once the signal is below threshold you specify).
9) some cheap dual-band APs have only single radio, capable of transmitting at both frequencies. Not at the same time, obviously, they time share. Make sure your APs have separate radios for each band they support.
10) some APs do support roaming (802.11-k/r/v); for home (or WPAx-Personal) that...
If you run a large scale Plex server I could see the 25Gbps coming in handy. Most 4K movies (Native Blu-ray rips, not the low bitrate renditions you get on streaming services) run about 50-100Mbps. That would let you stream to 200+ people at once, and have overhead for other things.
> I think beyond 1gbit, the benefit become super marginal and the hardware expensive
Is it expensive though? We spend how much of our lives online. How much do you think this all-spanning life upgrade costs? What is the price of never ever dealing with buffer-bloat? Take a guess.
I calculate it as a one time $356 cost plus labor. You might have paid more for your wifi system. 128 port 25gbe switches are around $20k ($156/port). Transcievers are under $100 and you need one on both ends. For a lot of already deployed fiber, this is a drop in replacement. This is absurdly cheap. Given how cheap this is it's an obvious & enthusiastic heck yes. Who wouldnt throw down $356 right now to get 25Gbe for life?
Cost gets a bit more complicated when it comes to the POPs & their uplink. Subscribers are going to be way oversubscribed even with some fairly expensive 100Gb uplinks. As you get further from an exchange the difficulty grows geometrically (because pops become.further hops away from the ix). Peering needs to be bigger too, as does transit (but ask whether the net volume of traffic grows elastically or not), which has costs. But I think we need to frame this question a bit better, of whether it's "worth" upgrading. Honestly costs are so low it doesnt make sense not to; the rest of the world is just milking us, bilking us, charging what the market will bear, protecting it's profit centers, and this company init7 is doing what makes financial sense for the consumer. Donwe need all that? Maybe maybe not. Should we settle for less? There's almost no financial case when the hardware is so so so very cheap. This tech sounds magical but 25Gbe is not exotic, not extreme technology; "the future is already here, it's just not evenly distributed yet".
It sounds like you're just taking about cost for the ISP to upgrade.
To actually realize your faster speeds, you need to spend thousands of dollars yourself on new switches and NICs. And then, as mentioned, the benefits are marginal. You would have to be streaming 10+ 4K movies at once to even "need" gigabit, let alone 25Gbps.
Wifi 7 is expected to be capable of 30-40Gbps. A dual port nic can be had under $200. Currently low/medium port count equipment has no demand, but perhaps the wifi7 world or pressure like init7 generates can make more visible & obvious the market demand. For anyone setting up today, do what I did: (used byt plentifully available) 18 port 40Gbps infiniband switch for $150, nics for $100.
I semi agree that I dont think we know what this is for. Never ever having buffer bloat is a tempting first ask. Connectivity is more than the sum of throughputs, as your figures imply- there's questions of availability too.
Being able to access each other's systems at near local speeds sounds quite compelling, could help jumpstart post-Big Social computing. You talk about netflix streams, but those are heavily compressed with the best offline encoding on the planet: if i just want to open Steam Remote Play Together & share realtime 4K with a friend, I'd need a lot more throughput since I have much much much less efficiemt encoding. If i wanted Remote Play Together with 3 friends, well, that figures goes up. If my family member also wants to do the same, now we're using a lot or maybe all the throughput & we're starting to have some contested bandwidth, some rising latencies.
The truth is somewhere between. Rationalizing ourselves down to what sounds sensible today, to me, is a cruel trick, is not just path dependency but an ideology that believes only in what we have & can see now, & refuses exploration & trying. To me the world & tech is spiritually fueled by why not thinking, by deciding to opt for the extra thats within reach.
Forgoing a cheap (still less than the price of a nice tv, by far), available one-time purchase option that vaults us into near-local connevtivity caliber with the world is still a lock in my book.
A router with a 25G uplink and a bunch of 10G sfp slots will set you back $600
There's benefits to 25G (certainly when transporting 4K video around which needs more than a 10G nic), whether that's worthwhile for a typical home is likely "no", so unless you've got hundereds of employees in an office it doesn't feel very useful.
A significant part of my life is online, indeed, but 99% of the time, bandwidth is a nonissue. Latency, very important. Bandwidth, not so much. I haven't experienced any bandwidth related internet problem in the past ten years, and that's moving from apartment to apartment, from hotel to hotel, from airport to airport.
Don't get me wrong. I'm very much in favor of this upgrade. I just don't think it's going to be an everyday quality-of-life improvement for most people. It's more about providing a service for people with special needs, and future-proofing the infrastructure.
> 99% of the time, bandwidth is a nonissue. Latency, very important
Indeed. And I'm speaking as someone who downloads most visual media before I watch it, so bandwidth matters to me. But not that much. With 100-200 Mbit/s I am good.
Cookie prompts, newsletter pop-ups, scrolljacking and ads constitutes the majority of wasted time for me, by a long shot. Latency to sites in other parts of the world can cause problems sometimes, since number of round trips can be quite high with TLS neg + progressively loaded content.
The calculation works only if each customer has their own fiber.
Unfortunately for most (consumer) FTTH deployments, that's not the case. Most of them are GPON, where the initial deployment was more effective, as up to 64 of your customers share single fiber, but then that means all of them have to upgrade all 64 of them at the same time, you cannot do them one at the time (see also the speed of XG-PON upgrades).
Additionally, many providers forced use of their CPEs. If you can send out SFP module and the customers can put it into whatever they want, it is much simpler, as replacing CPEs for all the customers on that fiber.
> Very few servers will allow you a full 25gbit download
As TFA mentions, same was the case with 1 gbit. I can confirm, I'm on the same ISP and was "early" to have 1gbit/s. Lots of servers still only had 100mbit/s links. These days I have zero issues saturating my 1gbit link.
Give it time, now that it starts rolling out, costs will come down and server links will be upgraded. The usual early adopter stuff.
This +10. For the vast majority of users, having 25Gbit/sec home fiber is like having a 250 MPH-capable supercar. Big, bad bragging rights. Token-at-best use cases.
5 people in my house, each using 5Gbit/s, sounds far more like the grandkids playing a "mine's faster than yours" game than like any "real world" usage scenario. Similar for 10 people each using 2.5Gbit/sec, or 25 people each using 1Gbit/sec, or ...
It's not 1998 any more, you've got multiple TCP streams, things like QUIC and other UDP based protocols, and of course good old fashioned window scaling which will go upto an 8gbit window, so as long as you've got an rtt under 300ms you'll be fine at 25gbit.
You'll encounter other limitations, like processing 2 million 1500 byte packets a second, which is possible but you need to make sure you have the grunt to process and the right types of options in the kernel
You forgetting the case where one hosts their own servers in the basement. Assuming that the connection is symmetric and offers static IP (which it does in my case but "only" for 1GBit/s).
Some time ago PC has revolutionized the world by giving access to computer power to general population. This has unleashed a tidal wave of creativity and business.
Giving the ability to host own servers to everyone can open up endless opportunities as well.
I host some of my own servers and benefit from it greatly.
Yeah, I think more home connections going symmetric could open up big personal computing applications that haven’t been super practical before, and make a market for easy-to-use personal home servers. Even just personal media streaming while traveling can benefit greatly from upload being boosted past the 10mbit up that’s somewhat common among cable ISPs. Having a symmetric connection also removes one of the mental constraints that we have at the back of our minds.
It’s my biggest hope for internet re-decentralization.
I have the same internet provider and package. While 25Gbps is indeed basically unattainable to anything other than Init7's speedtest server, it's easy to exceed 1Gbps.
Software and driver updates exceed 1Gbps all the time, as do game updates/downloads through Steam.
Piracy also works really well. Downloading copyrighted media is perfectly legal in Switzerland and I was able to get ~7Gbps real-world speeds from Usenet without too much hassle.
It's also really handy for things like backups. As of writing, bandwidth to a Hetzner cloud server is ~5Gbps up/down with iPerf3.
I do agree that 25Gbps is more overkill/bragging rights than it is real utility but I think 10Gbps is an easy sell.
Keep in mind that there is no change at all to the monthly price of your internet by choosing these higher speeds. 10Gbps has the same monthly and setup costs as 1Gbps, you only have to pay a bit extra for hardware capable of dealing with 10Gbps, which is pretty affordable.
Also with init7, very happy with my 1Gbps line. I'd probably even take a 300Mbps if it was half the price. Even with gigabit LAN I don't saturate the line more than a few seconds a day on things like updates, and anything from my RPi NAS isn't going to saturate it anyways.
Please note too, piracy is not legal in Switzerland. What is illegal is the spying and tracking which would be necessary to build a case and prosecute pirates. Technically you could still self-incriminate if you documented all of your piracy, with proof, and published it online.
Downloading is legal in Switzerland. For a nice authoritative source, here's the Swiss Federal Institute for Intellectual Property (part of the Department of Justice and Police)[0]: "Downloading copyright-protected works for private use is permitted in Switzerland (Art. 19 CopA)."
The mentioned legal basis is [1].
Note that this applies strictly to downloading though. Participating in a torrent swarm (where uploading is also happening) is not permitted. That's where the technicality you mention comes in: until recently, it was illegal to monitor internet users for copyright enforcement purposes, which meant it was illegal to monitor a torrent swarm, which meant you could somewhat safely seed torrents, despite it being illegal.
> But with 25Gbps you can just have multiple connections open without any of them ever affecting any other.
Realistically, this is pretty close to true with 2gbps symmetric, too. My provider seems to give me 2.2gbps in practice.
PS5 downloads/updates are 500-600mbps in practice. Steam is 1.5gbps or so. Most other things-- streaming, video calling, etc, are under 40mbps. So, you know-- if I kick off a PS5 download, and a steam update, and my kids are streaming and video calling... And my machines are backing up to the cloud at 1.5gbps (other direction)... and I decide to do a big apt-get update on a machine, maybe my steam update completes a couple seconds later.
Of course, I want even faster... but I'm hard pressed to say what would be better.
Agreed. A 25gbps download will fill an 8TB HDD in 47 minutes, so how much are you going to actually utilize that full bandwidth? Not that any webhost will ever give you even remotely that much bandwidth. I'm on 1gbps fiber and I've never saturated my connected. That would require 40 simultaneous 4K streams going at the same time in my house.
I'm wondering: is my storage fast enough to write at 25 Gb/s?
SATA drives max at 6 Gb/s, and probably a bit worse than that in practice.
M.2 NVMe SSDs can reach 25+ Gb/s on sequential writes. Maybe I could overclock these to squeeze a bit more out too. Although, an 8TB M.2 drive will cost over $1k for now.
I suppose there's also RAID. Although, I'm not sure what other limits/gotchas I'd run into there.
On behalf of all tech-minded Americans, I would like to say... I hate you.
I just got a notice in the mail that my ISP is "upgrading" their network so now I can pay $200 USD/month to get a whopping 2gbps, which I actually thought was pretty amazing until I read your post. So, thanks.
In all seriousness, congrats! You made a good case for why one would need that much bandwidth. Also, we need to catch up here. :)
I’m a tech-minded American that pays $65 per month for a 1 Gbps up and down line that I am completely satisfied with.
Until recently I paid $35 per month for a 300 Mbps line because I couldn’t justify the jump to 1 Gbps.
Then I had to upload some local Hyper-V servers to AWS to convert them to AMIs and figured what the heck I’d upgrade the line.
I would have no interest in upgrading beyond 1 Gbps right now though. There are too many infrastructure components that need to be upgraded to attain that and I don’t have a use case for it.
USA requires more raw materials to provide services. United States is about 19 times bigger than Thailand.
Thailand is approximately 513,120 sq km, while United States is approximately 9,833,517 sq km, making United States 1,816% larger than Thailand.
And then there is the duopoly of ISPs where they carve out huge chunks and agree not to compete.
I live in a metro area in the US and thousands of homes in my community do not have land-based broadband options. The US incumbents have totally failed and it isn't because there's a lot of desert in the West and Alaska. I'm sick of this argument which doesn't explain why city dwellers in most places in America have the worst internet in the developed world.
Yup, I live in San Francisco proper, and my only choice is Comcast cable. Looks like the current promo pricing for 1200Mbps is around $70/mo, but I can't quickly find what the normal price is. And I assume the uplink is something abysmal like 25Mbps.
(I'm on Comcast's Business service, $250/mo for 1000/35 [long dumb story why]. Most of the time I see under 600 down when checking on speed test sites, and real-world speeds downloading large files rarely exceeds 250. I expect real-world speeds on the non-business service are even worse.)
It's pretty embarrassing that this is the state of things.
Sheer size isn't what you should compare. Population density is much more relevant. Granted, Thailand still comes out ahead, but by far less than your size comparison (33.6/km^2 vs 132/km^2)
pop = {'USA': 329,000,000, 'THA': 70,000,000} #as of 2021
gnp['USA'] // pop['USA']
65805
gnp['THA'] // pop['THA']
7027
Seems more like a distribution of resources issue... so why not jack up taxes on the wealthiest 1% and use it to pay for things like high-speed fiber in all the rural areas? Doubtless this would lead to economic growth?
FDR's Rural Electricity Cooperatives did a lot to electrify much of the midwest and rural south, along with the creation of the TVA. I doubt anyone would want to rely on the current major providers, Comcast etc., who have such a bad track record, to accomplish this. Municipal broadband sounds like a better option:
> "The Rural Electrification Act of 1936, enacted on May 20, 1936, provided federal loans for the installation of electrical distribution systems to serve isolated rural areas of the United States. The funding was channeled through cooperative electric power companies, hundreds of which still exist today. (wiki)"
There are fair few places in the US where the local power company also owns a fiber network and provides (relatively speaking) super cheap gigabit or multi-gigabit internet service
However there are just as many places where the state's government was bought off to ban such networks because the majors are afraid of actual competition.
Don’t forget the existing providers have already received massive funds to ‘improve rural broadband’ in the same vein as that act. Hundred of billions of dollars if I remember correctly.
But is that within the city / country, or to outside the country, too? International bandwidth can be completely mediocre compared to local bandwidth. What I do is change the Server in speedtest.net to one in New York / wherever I have business in and test it out (from Hong Kong).
Where do you spend 120 baht a meal? That's a restaurant meal, not the average meal. I don't spend more than 60 baht in Bangkok. At $7 (250 baht) you're talking about a night out (maybe except alcohol), way off of the "average meal".
I love the tightest lifestyle eating rice and 1-2 side dishes from street vendors who deserve the money rather than foreign investors to extract money from the economy like leeches.
I worked in BKK since 2016 and office staff usually goes to the cheapo restaurants (which I frequented too), so 40-80 baht is completely common also for white collar workers. But meat, vegetable and oil quality is really shady in that price range.
I prefer to order grab of around 120-200 baht from a "healthy" place or going for restaurants within Central 200-400 baht per meal.
The funny thing about rice with something, most of the time you want to add an fried egg too (more likely like + 10 bath charge). Crispy Pork is about 10 bath more expensive than the other. I think 40 bath is like 5 years ago pricing. You can still get it somewhere. 50-60 bath is more standard now.
The US has around ~10-12 times the disposable median income of Thailand (2019 figures), and ~10 times the GDP per capita of Thailand (2022 IMF estimates).
It's like a typical consumer in the US spending $500+ per month for broadband. Even worse if you consider the rural income factor. Absolutely insane.
For a fraction of that you can get 1gbps from Comcast and you'll never utilize most of it in 99% of consumer situations.
I'm in a small quasi rural 'city' in the US, hours away from any consequential city, and I can get 1.2gbps from Comcast for 15-20% of the income adjusted rate in question referenced for Thailand.
It seems common to forget how astoundingly high US median income, disposable income, and GDP per capita figures are compared to the rest of the world. The latest 2022 estimates are pegging US GDP per capita at nearly double that of France and Japan. To match up on median disposable income figures, you have to use hyper affluent countries like Australia, Denmark, Sweden, Switzerland and Norway as references. Then people come on HN and proclaim how they're paying only $20 per month for Internet access, in a country with 1/10-1/5 the median disposable income of the US. The US is more expensive than it should be for Internet access (better telecom competition would go a long ways toward fixing that), however the reality is US income figures are also a lot higher than most of the developed world.
I’m just wondering how good of a measure it is comparing the median incomes of different countries. If a there is a country A with 10x median income of some other country B , then the people of country A are really 10x more affluent than that of country B? especially if they pay 10x for everything?
That’s not a reasonable comparison for Thailand food prices as food isn’t disposable income. World bank says GDP per capita PPP was 18,232$ in 2020 down from 19,233$ in 2019. Of course that’s not evenly distributed, but rural vs urban incomes mean costs are higher in cities than median income suggests. https://data.worldbank.org/indicator/NY.GDP.PCAP.PP.CD?locat...
Where your number comes into play is for people pricing Netflix subscriptions.
GDP PPP per capita is an extraordinarily low quality metric.
You end up with absurd examples where Botswana is comparable to China; Russia is comparable to Greece; Puerto Rico is comparable to Spain, ahead of Portugal, and just a bit lower than Japan; Kazakhstan is just a bit behind Latvia and Slovakia; Taiwan is far ahead of Finland, France, UK, New Zealand.
It’s just a question of what you’re trying to use these numbers for.
Imputed rent for example is one of those things that’s kind of silly on the face of it but makes various comparisons more reasonable. On the other hand it can also imply a great deal of economic activity that isn’t actually happening.
PPP is the same sort of calculation. If rents crash because a great deal of housing was built it can make GDP comparisons kind of meaningless. The country has more tangible wealth, people are better off, yet GDP falls. That’s not what you want the number to represent.
Both Nominal and PPP are out of whack if you are looking to learn about conditions on the ground. GDP measures the production of a certain country (and it's a very bad metric at that). Some countries are wealthy because its people make money from foreign sources. This is usually displayed by a high and chronic trade deficit.
That means you can have two countries with comparable GDP per-capita, where one of them have a more affluent population and able to pay higher prices.
I'm in a top 10 most expensive city in the world (Canada) and the best I can get for $50 is 80 Mbps (and even that is only available as a promotion that you can't get by simply going to the ISP's website and buying a plan).
I've looked at the smaller ISPs in my area and they're terrible as far as pricing goes. Both the "major" options are still using Rogers' last-mile infrastructure, they offer poorer customer service (because of the previous point), charge the same or more money for the same level of service and offer no real incentive to switch. Having talked to some techs at TekSavvy, none of the smaller ISPs can offer anything interesting like synchronous speeds over coax/DOCSIS because they have little/no control over how the last-mile infrastructure is run. For that same reason they can't offer anything faster than 1G down either. It all feels like smoke and mirrors, and the CRTC seems to have a vested interest in keeping internet prices sky-high.
Yeah the last mile stuff is still owned by the majors, but at least the choice exists.
As much as Canadian telecom seems to own the CRTC it isn’t nearly as consumer unfriendly as when I lived in the US.
I had an apartment in downtown Seattle which had 2 choices for internet. Many of my friends were quite impressed since most addresses are only served by one ISP.
I'm in Vancouver and paying about 55 CAD plus taxes for 300Mbps. So it's possible to get a bit more for similar money. The downside of that is however those are not available as regular offers, and you constantly have to deal with ISPs and rention programs to keep the price down. Even had various events where the ISP randomly increased the price inside one month, until I gave them a call and ask to fix it again.
This price randomness never occured to me in germany, and just booking a fixed low price on a website was so much more convenient.
I'm in Vancouver, BC and paying 80$/mo for 1Gbps symmetric. I could get 2.5Gbps but it would be about twice the cost - and I would have to get something that can do 2.5G link speed on an SFP+, not many devices can.
With a best effort connection Id rather have a well oiled setup, without cgnat, a good cpe, sensible throttling etc and 100mbps than whatever huge amount of gbps.
I don't think they do any longer but Verizon used to send me FIOS flyers on a fairly regular basis and when I went to their site it looked as if it wasn't actually available at my address (where I get Comcast).
Also amusingly, both Comcast and Verizon have the very old (as in multiple decades) name for the street I live on as my address. It's only somewhat wrong (the name used to have a North on it)--but it is still wrong.
In a large French city with 10gb/s fibre connection at 50 euros à month (internet, TV, fixed line phone). More than enough for two people working from home + family.
That’s all very well if you have a tech job. If you have a low paid job you’re utterly screwed in the US. I’d rather make less so my fellow citizens can live a decent life too.
Brooklyn here, my building still doesn’t have fiber — I’ve complained to the city for years! I get to pay $75/month for 200Mbps down, 10 Mbps up. I’ve been told fiber is coming for almost 15 years.
Reading this thread, I am reminded of how cheap telco prices are in France: e.g. in 2020, fibre internet plans averaged between €26 and €28.35 (USD 28-30 at current rates) per month.[0]
Mobile plans are also on the cheap side, to the point of being competitive with many third world countries (for example, I wasn't that impressed with prices in Thailand in 2020 for comparable plans, only a 1-2€ difference with what I had at home).
This is getting to the point that I find myself often suggesting fellows from border countries open a line in France, just so that they can enjoy the "European roaming data envelope" that comes with most plans (i.e. several gigabytes are free to use from anywhere in Europe, and you can make calls and texts to other EU countries when abroad...) after realizing how expensive and suckish mobile plans are in their country (Belgium in this instance).
Which provider would you recommend? I remember taking a look a while back and the most viable option was Free, but 15.99 EUR a month for when I need roaming 3–4 times a year seemed a bit excessive at 180.- per year.
For your specific use case, I'd say give Lebara mobile a shot.[0]
SIM card is free, you can top it up with a prepaid amount or do a one-month plan without renewing it, and I just checked their T&Cs relating to roaming and they are actually quite generous[1][2].
Besides, their website is (for the most part) translated in English due to their target demographic, so no need for full French fluency here!
Some good info in English on the French telco landscape can be found at [3].
I'll nention it for completeness sake: another possible solution, albeit a bit cumbersome, would be to get a regular contract-free plan but suspend it for the period you don't need it.
A cursory search indicates that both Orange and Bouygues (as well as their contract-free offshoots, Sosh and B&YOU respectively) still offer this as a paid option at 5€ and 7€ per month respectively.
This means you could get the plan appropriate for your needs, then suspend it (for at most 6 month per year it seems) and only reactivate it when you need it.
However since this procedure requires getting in touch with customer service, it should be reserved for people well versed in French :-)
I was under the impression that they only requested an ID and a mean of payment, but after verification you are right, a proof of residency is also required for postpaid plans.
If you know someone in France, you may ask them to make a "solemn declaration" that you live at their address; then you only need provide their own proof of residency (people living with their parents routinely do so).
Tourist and prepaid SIM cards are still exempted and, depending on your case, can prove quite interesting.
There are other ISPs in the states that are much more competitive (I work at one). Unfortunately, if you're not in a fairly dense area, the chances they can deliver to you is minimal, since building your own network is expensive.
For example, we offer $40 service, and if you're in our historical areas that's 1Gbit symmetrical, and if you're in newer areas we've turned up in the last year that's 10Gbit symmetrical.
We're expanding (as I expect most ISPs that can undercut the major players that much are), but it requires actually stringing fiber through neighborhoods on poles, so it takes a while (but that can be scaled...).
If you don't have telephone poles, it's much harder/more expensive to build out an area, so often those are skipped (at least initially) as areas cheaper to deliver to are prioritized. The is unfortunately a lot of new development, as they'll build neighborhood with underground utilities and pre-wire AT&T and Comcast, making it hard for others to deliver to the area without a lot of cost and work (trenching).
I have a sometimes 12mbit connection, but there's a Fiber POP about 1 mile down the road. There are phone poles (with fiber on them) between here and there. What does it cost to string fiber on poles, and what are the hopes of getting right of way? Alternatively, is there some way to pay / force the telco to build out FTTH?
Some people a few miles away have a fiber POP at the end of their shared driveway, and are trying to decide if they should pay the extra couple hundred per house to go from 1GBit to 25GBit symmetric to the houses.
In the US I believe anyone can use the pole space as long as it's not taken, but I'm not an expert. I do know you have to fill out engineering documents per-pole to explain the load and propose (pay for?) fixing the load bearing attached cables.
At the ISP I work for, I believe we spec out the cabling we need to a neighborhood and then order a special bundle with breakouts at specific locations along the length to serve locations, and then string that along the poles. I'm not sure why we wouldn't serve a house with one of those, but those go back to a central point in a neighborhood, and it's possible the backhaul from the central office to that central point in the neighborhood passes houses that aren't served. Where to build is all about ease of wiring an area and housi g density. It's all about cost per houses passed. The good news is that maybe your area is slated to get fiber since it goes by there, and it's just a matter of the lower hanging fruit being picked first.
Thanks. The polls are ancient (1980's; most bow 10-20 degress under the tension of the existing tensioning cables.)
This street got sold off to a bankrupt telco (from what I can tell, their business model is to buy up unprofitable lines, then periodically file for bankruptcy).
There are a few local fiber providers. Maybe I could call and see what it would take to convince them to build out. (There is definitely demand here.)
Alternatively AT&T (who is not our local telco) keeps spamming us to get a business fiber to the home connection. They make it clear they'll ban customers that appear to be using it for residential traffic, but I think that sort of traffic discrimination falls afoul of California's network neutrality laws. Has anyone tried calling their bluff on that?
It's a lot of work, but if you survey the area and get people to sign something saying they would buy pay up to $X for gigabit fiber or better, you might be able to approach some of those independent ISPs and lay the case out, you might get them to reprioritize your area (but it may or may not cover the same area you surveyed).
Be careful approaching incumbent major ISPs or letting them know if you got the signatures though. You might find they went out and offered sweetheart deals to the same people on two year contracts, but at existing service levels...
Regarding business fiber, I think they're more trying to keep people from reselling it as a micro ISP. If you're willing to pay business costs for personal use, I don't think they'll care. If you split the cost between 10 neighbors, they might. There may be other concerns they are dealing with that I'm not thinking of though.
AT&T called me. They said their technical guys think they might be able to upgrade the connection to my building from 1.5Mbps to 3Mbps if I want to sign-up with them. I'm in downtown Chicago.
The guy on the phone was so cool. He was totally "in" on how hilariously awful it was that they could only get 1.5Mbps into my building. He had access to the system that let him see all the contracts of my neighbors and what connections they had and was floored that every building next me had gigabit.
One thing most people ignore: the prices these days for many thing don't have a direct correlation with the cost, but with how much people are willing to pay. This is happen especially in quasi-monopoly situations, like housing and internet services.
For example I have 3 Internet connections at home that I pay ~ $30 in total, from 150 Mbps to 1 Gbps. I think the 1 Gbps is ~ $12/month, it is so cheap because there are so many options and I can afford to keep all 3 for redundancy (1 is a 4G mobile data capped at 200GB/month, it works even when there is a power failure in the entire neighborhood).
But if there is no variety of options and no competition, then price is high. From what I read, most ISP in USA are squeezing as much as they can from their clients, sometimes to ridiculous levels.
Here in Australia: $90/month nets you “100/40mb”, which is really more like “~85mb down, 25 up but only a quiet night, if you’re next to the router, and it hasn’t rained within the last week”.
It’s much, much better than it used to be at least (depending on where you live + NBN technology choice of course). And yet still completely hamstrung by the mess the current government made of the original NBN plan.
If you live in an area that can get HFC, FTTP, FTTN etc. it’s possible these days to get a pretty consistent 1000/50 for $150/month. Those fibre services are pretty stable in my experience, definitely a million times better than the old broadband which as you put got destroyed by bad weather and distance from the exchange.
The point is, that even if I were to plug my computer directly into the router, I'm not getting _good_ speeds, I simply wanted to head off comments like "have you tried 5Ghz networks? Have you tried a quieter band? Have you tried x/y/z other thing?".
I'm currently on 1000/50 cable internet, which is already quite nice. Telekom is laying fibre but it's not clear yet whether or not they will stop short of this house. Also i suspect it will be a while before they offer better than 1000/500 service.
I hate asymmetric connections. You get gigabit download but something stupid like 20-100 Mbps upload? Just why... Upload speed matters just as much as download these days.
Because you have a single copper pair/coax cable with a finite bandwidth on it, and making a choice about how to split it.
Eg, your tech and cable can support 40 Mbps. If you split that 20/20, your users will have trouble watching 4K video that wants 25 Mbps. Change that to 35/5, and for most people it'll actually work better.
This issue goes away eventually as you either get a fiber for each direction, or you can just push terabits through fiber anyway, so there's plenty capacity to be symmetric without any compromises.
Well, that exceeds my knowledge, but my guess is that it'd be very tricky to accomplish.
You have one cable, a given frequency may go in one direction or in another. Both sides have to agree on what it's being used for. You'd need a communication channel between the ISP side and the client side to constantly negotiate, and that negotiation would take some time, so such a mechanism would have some latency to it, with possibly weird effects on things like online games. You could get weird behaviors where some particular pattern of traffic would result in the connection readjusting itself just wrong on a regular basis and result in hordes of angry gamers.
I think it's reasonable to guess that ISPs targeting consumers have no interest in monitoring and troubleshooting such a thing when they can just set a fixed split and be done with it (and ask the customer to upgrade to a bigger plan if it's not good enough), and ISPs targeting enterprises have no need for it.
Edit: And there's the issue of how you sell such a thing. It's a system that readjusts itself automatically based on some arcane magic and may work differently from one day to another. How do you make any reliable promises about it?
Oh yeah, missed the "cable" part. But I've seen the same thing with fiber in the UK and Germany. So the ISPs upgraded to fiber but kept the old (very) asymmetric speeds... I mean, at least have it be something more reasonable like 1000/500 :)
But I'm spoiled, growing up the only ISP in town spoonfed everyone higher and higher speeds even though no one asked for it. They're laying fiber to villages seemingly just 'cause. People are choosing 4G over fiber because it's cheaper (even though data is limited), go figure.
Well, actually there was a bit of government initiative (with no funding or enforcement) to boost the IT sector. The US would benefit much more heh.
What really sucks is the ratios these cable companies go with, 1000/50 is a 20/1 ratio. Having something like 500/100 would be light-years better for the vast majority of people even though you would be sacrificing overall bandwidth.
Perhaps it's a conspiracy to have less tech-savvy work-from-homer's keep bumping up their speed plan because their upload speed isn't up to snuff for video meetings. Either that or cable companies feel they need to compete on advertised download speeds with fiber, which is a losing battle.
In Zürich city a law was passed to pull fiber into every home. The fiber is serviced by the electric and partially the largest phone company. Any provider can use the network and therefore the fees a low.
If you live outside the city you may not be so lucky and spend 100+ on maybe 500mbit internet.
Init7 is also currently in a legal fight with swisscom because swisscom wants (has already started) to pull some alternative fiber that doesn't directly connect the customer to the pop therefore preventing competitors offering faster service than swisscom. The small cost savings that swisscom has is a huge problem for future upgrades and requires power in underground shafts. Not very green when you could pull the fibers directly.
I just upgraded to 10GB networking in my house and the old desktop I have hanging around mostly for guest use tops out at around 6 Gbps. The CPU just can't handle more than that. Granted, it's a 10 year old, cheap CPU with integrated video, so it was never much good. But still, it works just fine for web browsing.
And of course hard disks get 200MB/s on the very rare occasion, and often a good amount less than that. Even SSDs are limited by SATA's maximum of 6 Gbps, so make a plan for upgrading everything to NVMEs.
Damn, that's impressive. Technically I could get 2Gbps with the two providers running their own fiber in the neighborhood. But I don't even fully use one gigabit connection, running a torrent client is about the only thing that can do it. Any ideas welcome :D
Curious thing, the ISPs don't oversell even though they easily could. I get exactly what I pay for, speed is never below ~940Mbps (down or up), and uptime has been stellar.
Yeah; there really aren't any great use cases I can think of for more than 1 gbps to the home at the moment. A 4k stream runs ~25 mbps, a 100 GB game download will tend to cap out before saturating the connection and still take ~15 minutes (hardly a wasted evening if you have to wait for it). At 2 Gbps, if you actually saturated the connection downloading something, you'd better have an SDD, since you'll be downloading faster than a HDD can write. More bandwidth may be -nice-, for the few services that can take advantage of it (I'm not sure PSN, Steam, etc, even will), but it's hardly a game changer for home use. Even the backup/transfer cases he lists I'd hope are transferring diffs, which would likely make the speed increase unnoticeable.
But, of course, I'd still love 25 gbps fiber to the home just because.
Back when covid19 started, we all went home, the office closed, and then we had a server failure. And I spent multiple days downloading stuff to back it up before reinstalling the machine, because there was not enough spare storage inside the office.
If you work with VM disk images even 1 Gbps starts feeling a tad sluggish.
What about remote work? Why the heck are you downloading stuff to hardware located -at your house-, and not a datacenter, another office, the cloud (someone else's datacenter :P), etc?
I work with VMs. Disk images stay in the datacenter, and occasionally get sent between datacenters. They don't live at my house.
My experience with fast internet has been that most CDNs are just not routed well and are quite slow (especially Microsoft). Only a few good ones allow me to saturate my Gbit internet. Not sure what I would do with 25 Gbit though.
What I do enjoy is that rsync.net is using the same ISP so I can max out my upload to them.
Many servers specifically throttle single connections, which is why there's stuff like JDownloader and browser add-ons that download the same file in several parts over multiple connections
I'm really unsure as to whether it's worth investing whatsoever in consumer grade fiber connections.
My speeds are advertised as 1Gbits down and 100Mbits up. On speed tests, I get much higher downloads (~1.5Gbits per second). This might be because it's not capped properly, and I live in a place that they're still building houses in (I imagine that our "box" is not yet saturated).
However, I rarely max that speed out in any download. Video games on Steam max out at maybe 65 megabytes per second. wget commands go from between 10 megabytes per second to maybe 40 max.
The thing is, I know that the speed tests are legit, because I can do several of these things at once and none of them lose any speed.
Yeah, maybe. My connectivity feels pretty intermittent to some ASNs, but I don't know if that's what changes between a good peering connection and a bad one.
Microsoft’s CDN is awful. I’m in Chicago so my fiber connection literally terminates at an Akamai and Cloudflare edge. Steam, Apple, et al I routinely pull 100+ megabytes per second. Microsoft, I’m lucky if I get a third of that.
I think you can now put 800 gigabit per second today, via commercially available kit, down a single strand of fibre installed back in the early 1980s.
Copper or radio connections just don’t seem like as worthwhile an investment to me. I’m not saying it makes sense to have an 800gbit connection to a residential property but from a longevity, cost per mile, use of less rare materials, reduced rf interference or whatever, fibre just seems a better proposition to me.
That’s the beauty of single mode fiber, you can use the same pair for something as slow as 10 Mbps to 800 Gbps. Makes a lot more sense from a data center perspective too, since you can continue to reuse the existing cable plant and just need to upgrade your NICs, transceivers, and switches.
The current commercial state-of-the-art is quite a bit higher, 25.6Tbps or thereabouts using a full 40 channel/color DWDM and 400g modules. But yes, copper or RF is IMO a non-starter for any new deployments.
I had 10gb internet on what was essentially a shared connection when I lived in Zürich.
That, along with a surplus server I literally housed in my shoe closet, gave me the firepower I needed to prototype out something that led to two research grants which now employ myself and a new PhD student.
We discount technological investments like this as being “too much” and “unpractical”, but we forget that, even if it only one person in a hundred or a thousand take advantage of them, the impact can be enough to launch careers or start businesses with sizeable positive externalities.
Being from a third world country, I completely agree with you on how technological investment can change lives. Even a low speed connection in a rural zone can achieve wonderful things.
But I really wonder what was your prototype that needed that speed connection and that wouldn't work with something slower. Can you share, at least vaguely, what was your work?
> Can you share, at least vaguely, what was your work?
We work with gis/satellite data. Nothing groundbreaking technically, but we aggregate it into products that are useful for social science researchers who aren’t comfortable with or don’t have the computing resources to use this data by themselves.
I can think of two things. Large time series data for large areas with a lot of attributes like temperature or other atmospheric data or a huge amount of point clouds or comparable 3D Data.
SAR and multispectral imagery. It’s not difficult to work with when you’re only dealing a small and well defined regions, but the bandwidth and storage requirement definitely increase once you start doing daily global composites.
We’ve actually moved everything into a data center with a 1gb connection. The trade of being that we have several orders of magnitude more storage and computing capacity.
It’s crazy because among friends I’m the only one working with geodata (remote sensing, too) and the last few days I keep stumbling upon people who already do some work in the same area and are probably much more advanced than I am. Good luck with your endeavour!
I was surprised when my parents had 1Gbit in 2009.
I was delighted when I could get 500Mbit in 2013.
I'm slightly miffed that I don't even get 1Gbit in 2022.
I literally laugh at providers attempting to convince me to get whatever with 10 or 20Mbit uplink speeds, in 2022l Yes, they exist.
Anyhow. What's insane about a 10Gbit uplink? I wired up my apartment for 1Gbit in 2001. I've been frustrated about home network speeds for 21 years. I do not understand why you consider 10Gbit insane.
At home you are free to make everything 10G-capable, have a NAS that can saturate such a link, etc.
On the wider internet, few resources are happy to dedicate such bandwidth to you, even if you can consume it. Not even every popular torrent can muster so much incoming bandwidth.
I can see how it may make sense if you run a popular public server from your home; maybe more convenient than colocation, but still is not a typical use case.
I feel the same. I am an übergeek, but I see no reason to pay for more than 100Mbit/sec, but 1Gbit and (soon'ish) 10Gbit will be available. Where I live, 100Mbit/sec is laughably cheap. I get a sales call once per month: "100Mit/sec is so slow! Would you like a upgrade?" I have no use for it. That said, I view myself as the exception to the rule. Please don't read this post as "1Gbit+ is useless" -- only that I don't have a use.
Real question: If you are an Internet gamer, I assume that latency per packet is lower for 1Gbit+ connections, so it is useful for many customers.
More: I feel the same about monitor resolutions above 4K. Anything above 4K on a 27inch monitor makes no sense for my body size / sitting height. I cannot use a vertical monitor taller than 27inch, and above 4K is too small for my aging eyesight! Again: I am sure there are many interesting applications for 4K+ and 27inch+. I hereby wish good luck to all the whippersnappers who will prove me wrong with interesting applications of 4K+ and 27inch+!
You can look at it as too much for anyone's needs, or you can see it as enabling tomorrow's applications. Looking back, or looking forward.
Back in the dialup days, nobody in their right mind would run a torrent, or play online shooters, or watch streaming HD video -- such things were utterly impractical over ultra slow links with very high latency. Once people started getting broadband, these applications started appearing.
Of course we need faster links into people's homes. Of course under-1Gbit links into homes in 2022 are ridiculous.
I have 100up/50down (mbps, not giga) at home, and its the fastest I can afford. If money is not a concern then it can go up to 1gb, but not in my area yet. I can barely saturate my line unless I'm donwloading something big.
I have 1Gbit Virgin in the U.K. and that speed us still very rare here. Any faster and I’m going to have to be searching for a better USB NIC in order to actually utilise anything. Most consumer hardware seems to still be 1000baseT so how would you utilise anything more than a gigabit connection unless you have 10+devices all working at maximum capacity? Also I rarely max it downloading anything at all, is there some sort of special internet somewhere that you can download stuff at 10Gbps?
Finally, can your SSD even keep up with speeds like this?
**Oh and I thought I’d note, it is the positive use of “insane” like “mad good”.
> I literally laugh at providers attempting to convince me to get whatever with 10 or 20Mbit uplink speeds, in 2022l Yes, they exist.
Meanwhile here in Silicon Valley, the fastest connection I can get at home (sonic.net) is 22Mbps/1.7Mbps. In the office I have Comcast Business which is a bit faster (don't recall exactly) but certainly well below 100Mbps.
So yes 1Gbps sounds quite insane. 10Gbps is out of this universe.
25Gbps is not insane if you are using commodity hardware. If you are buying any sort of datacenter switch, most likely it will be 25G/100G already.
10G/40G is dead tech.
Now, 400G uplinks are still pretty reserved, but 100G (and its' breakout, 4 x 25G) is common as dirt now.
Are there any small (affordable, energy-efficient) switches with those speeds? Say, around 8-24 of the 25G lanes? You can buy a dual-port SFP28 card for 250 EUR by using the QNAP "accessory"[0] which is just a ConnectX-6 Lx EN with a fan to work in the NAS systems they market it for.
This is quite efficient due to the PCIe4 support not hogging many lanes (half of the 8 lanes can be enough depending on usage), but to use the speed one needs a 25G switch. Is the cheapest way just an Ryzen with 5 of these directly attached to the CPU, and running routing on the CPU via DPDK?
MikroTik is a market leader in barebones 10Gbit/sec routers. However, multi-port 25Gbit/sec is very expensive! This model is 2800 USD and has 12x 25Gbit/sec
and 2x 100Gbit/sec ports: https://mikrotik.com/product/ccr2216_1g_12xs_2xq
I'm sure there's >10Gbits of 'gis/satellite data' available every second if you want it?
Or what if you want a bunch of historical data for some sort of analysis, and you'd like to start that now, not next week when it finishes downloading? If it's historical (i.e. sitting there available to download) then faster is always better right? If your connection is the bottleneck then all else equal why wouldn't you want it?
(I imagine 10Gbps is about the point currently where it's not necessarily going to help going faster on your end, unless you you have multiple (concurrent) sources, or know that even at that you're still still the bottleneck.)
GIS and remote sensing has you working with gargantuan TIFF files and massive amounts of other kinds of data. It's a whole thing. If you're doing data aggregation, it's possible that without an absolute unit of a pipe just to do a download of imagery, the prototype wouldn't have been feasible.
> Being from a third world country, I completely agree with you on how technological investment can change lives.
Being from a soon north korea 2.0 to be (russia,) I can say it was my surprise to see Internet being so bad across the developed countries when I was travelling in the previous decade. Canada - ridiculously expensive traffic, Germany - no comments, UK - sometimes good, sometimes 128kbps DSL, USA - 20mbps DocSis everywhere
India, Pakistan, Bangladesh, China, much of Africa had FTTH as the dominant way of home Internet connection for a while.
As a German I weep and cry. 25Gbit/s seems so so far off. And I live in a major city. I only get 150Mbit VDSL at the moment. I have no cable connection so one of these to get the theoretical 1Gbit/s download is out of the question.
Gigabit internet is quite widely available through different cable providers in Germany nowadays. Also the country side seems to be moving up, the very rural place where I grew up (and where my mom still lives) had a max of 2Mbps DSL for the last 20(!) years and now the whole area is being upgraded to fibre and will enjoy 10/1Gbps by the end of the year!
I moved into a newly build house in 2014 and was shocked to learn that all the houses only had basic copper telephone lines and Sat-TV. The whole field was empty and they had to do the groundwork for the copper cables anyway. I was shocked when the Telekom person, who connected my then 16MBit/s ADSL contract I had to move with, told me that the next TAL (connection point; Teilnehmer Anschluss Leitung, I don’t know the correct English term) was 5km out and that I will only able to receive 10MBit/s max. Netflix HD was blurry and browsing while streaming impossible.
I hear news that it gets better and that rural places finally get faster speeds but as long as I live where I live now I’m bound to VDSL or find enough neighbors who would be willing to ship in to get a Fibre connection.
Just fyi as I know you aren’t a native speaker, it’s ‘chip in’, if you were native I’d assume a typo, probably is for you too, but it’s a phrase easy to mishear and when I was learning a second language I appreciated these corrections.
The drawback is that cable it is a shared medium, so it can be quite bad when demand is high (in the evening) and the upload bandwidth usually is very low.
Interesting. In populous areas of the US they use HFC so the cable to your house only services a few buildings, with the neighborhood having a fiber optic back-haul that is shared, but much faster
Lived in Germany for 5 years and cable internet was generally terrible. We had 200/20MBit. But the actual upstream would often be 1MBit. Downstream was better but at many times not great. There would also be regular outages, that would take hours to solve. The only alternative was VDSL with a maximum downstream of 50MBit.
We moved back to NL and have 1GBit fiber, and there has been a short outage once in three years. I know that there are a still a lot of addresses without fiber, but when I last checked the stats, about 50% of the addresses has the possibility to get a fiber subscription. Heck, even my parents who live in a small rural town have fiber.
Yeah the classic argument, but at some point every internet connection becomes a shared medium. It really depends on how the network is setup and where the fibre backhauls start. If the building has older wiring which can only support 1Gbps and you have a bunch of high bandwidth users, then yes it can affect your bandwidth more than using other technologies.
We have similar situation in Poland. I live in rural area, but quite close to bigger city and enjoy 1Gbps for the last 4 or 5 years.
I wonder how the upgrade might look considering that 10Gbps hardware is quite expensive (and house cabling might need upgrading) and 2.5Gbps/5Gbps is quite new and hard to find router or laptop dock/hub supporting it.
The maximum speed available at my London address is 35 mbps download and 5.5 mbps upload. It hasn't changed in the seven years I've been living here. The best mobile connection I get is two bars of 3G unless it rains.
It's not typical for London. Not that extreme anyway. I think our postcode is among 2% or 3% left without a fiber connection. Many others have 100 mbps or so (I try not to look what they really get :)
Are you with hyperoptic or community fibre? I’m with the former and see speeds around what you’re saying. I think CF are rolling out 3gbps connections in my area though.
My previous address was stuck at 8 mbps dowload and 0.25 mbps upload and it will not be upgraded anytime soon because every corner of that street is listed / protected. I literally moved just because of that. Not I won't rent anything without fiber.
We are still creating newbuilds in cities without fiber.
In other news, someone fucked up construction and left an entire complex of brand new apartment blocks with 7 mbps internet
A lot of it is down to those countries being early adopters of telecom infrastructure. It meant when a lot of other countries were wiring up internet for the first time, they weren’t faced with the prospect of having to piggyback existing and aged infrastructure. eg in some cases they could leapfrog copper almost entirely and jump straight to fibre.
>in some cases they could leapfrog copper almost entirely and jump straight to fibre.
I visited Nepal several years back. They were going straight to wireless. All of the villages in the mountains had a solar panel on their roof that they used to charge their cellphones and there was a cell tower on the ridges.
It was surreal to see villagers use Facebook but have no road access to the main city.
Politicians in "old world" countries explicitly decided to stick to existing copper infrastructure instead of rolling out something better. These countries like to pretend corruption isn't an issue for them, but then you have asinine decisions like this.
In Canada it is down to our telecom oligopoly, which our government protects by (a) refusing foreign competition and (b) installing industry heads to run the consumer protection regulator, i.e. allowing the oligopoly to capture the regulatory body.
In actual fact, our telcos were heavily subsidized during their formative years, granting them a monopoly, rights of way, and helping to pay for their infrastructure. In return for a guaranteed profit margin, we had extreme control over their pricing structure and guarantees of service quality and coverage.
Then we allowed them to be privatized and deregulated, in exchange for which we get fucked. Which is, as far as I’ve ever been able to tell, the inevitable outcome of converting public services to private.
Govt created monopolies are bad whether they are public or private. Private is worse because money is removed as profit, not just reinvested.
Privatisation is good if competition is fostered at the same time. For example by opening existing physical infrastructure to multiple service providers.
Where I live we had (historically) one fixed lined govt provider. When mobile arrived we got 3 private providers. With fibre we got multiple hardware installers (ie multiple companies digging fibre) and then dozens of ISPs running on that.
Same as banking. Inventors and early adopters get stuck with "works well enough" old systems and all their deficiencies and limitations. Late newcomers roll out newest and greatest solutions.
This is term is commonly misused as it originally referred to countries beyond the Cold War's NATO/USSR denominations. Today the parent likely refers to a country in the global south or a "developing" nation.
I'm from what used to be a third-world country in the strict sense of term, I'd now refer to my country as a country in the global south.
It isn't a geographical denomination, but rather socioeconomic and political.
Most of the countries in the global south denomination are actually in the northern hemisphere.
Is it south of some reference point (here in Maryland it looks like I’m south of Turkey and around the same as Beijing). Is Australia excluded? Or is it just a euphemism for developing nation.
There is not as it does not refer to a geographical designation. I wouldn't say it is a euphemism, but closer to what people actually refer to when they misuse the term "third-world".
The closest geographical designation was the Brandt line drawn in the 80s showing a north and south divide in terms of global wealth distribution [0].
Like I mentioned in my top-most comment, "developing nation" is what people typically refer to. As to your comment, I was pointing out that "global south" is not a euphemism for "developing nation". I'd say it is the other way around and in a misguided way. This is because the global south/north designation intentionally distances itself from terminology such as "developing nation".
I understand if this comes across as pedantry but the global north and south terminology does a better job at leveling the playing field when talking about wealth and progress disparities.
As per your other question, Australia is part of the global north, China is part of the global south in the global north and south groupings.
>I understand if this comes across as pedantry but the global north and south terminology does a better job at leveling the playing field when talking about wealth and progress disparities.
I mean... but how? It's just the exact same concepts with less intuitively obvious words. Arguably it frames the disparity as a more immutable part of a nation; it doesn't fit normal language patterns to talk about a country moving north/south.
> I understand if this comes across as pedantry but the global north and south terminology does a better job at leveling the playing field when talking about wealth and progress disparities.
How? South Korea, Japan, and Singapore are richer than France or Spain now. As nations develop the geographic terminology makes less and less sense. Developing nation is much more accurate.
> In other words, effectively developed = what we are, developing = trying to be what we are.
That’s the plan, yes.
> Ultimately it is forcing a finite game onto the world.
Global south sounds like a better term.
Except that it’s geographically inaccurate as well as kind of racist. Japan, Singapore, and Korea can get richer than France or Spain, but they still get lumped in with the other non-white people in “the global south.”
> Japan, Singapore, and Korea can get richer than France or Spain, but they still get lumped in with the other non-white people in “the global south.”
Global south/north divide lumps SK, SG and JP into the north. I have a bigger problem with it putting Russia there, but not e.g. Argentina.
Besides, note how the term "developed" sounds like it has an objective, final and fixed definition, but always invites moving goalposts. Now you use some financial metric. Tomorrow someone uses education levels or democracy.
**
The truth is that everybody should keep developing, and development can entail different things to different people. If you say "I'm developed, and that other person is developing", "we are good, but they are not good yet", all you get is resentment when in fact you want global cooperation.
"Developedness" is a vague, subjective measure that indicates duck all except hubris on the part of whoever coined the term (not coincidentally, that was developed countries).
A vague measure warrants a vague term. "Global South" is far from perfect, but it is less bad. It is obviously geographically inaccurate, which on a meta level is quite fitting.
I don't think I could ever use either term without air quotes, but at least with global south they are more or less implied.
> Global south/north divide lumps SK, SG and JP into the north.
Only if you arbitrarily make exceptions for what’s “south”—ignoring the fact that Singapore and Taiwan and Korea are south of China. The phrase itself implies a permanent association between “south” (I.e. non-white, since skin color is a function of latitude) and “under-undeveloped.”
That seems far worse to me than acknowledging that developing countries are indeed trying to become like developed countries.
Korea is hardly south of China. Regardless, the term is not supposed to be geographically correct, and to me it is a feature. It's a vague subjective term, as if it's been designed for people to think twice before using it and make their point using a more precise metric instead. That's why I think it's better than "developing nation", which is just as arbitrary but tries to hide that.
What is the measure of richer? By median income, Japan and South Korea are slightly richer than Spain but poorer than France. They are higher by GDP per Capita, but I don't think that is a very good measure as it completely ignores extreme income inequality and would list several oil rich nations very high even though only a few very wealthy at the top benefit.
It isn't as much as a replacement term, but you make a fair point that is doesn't translate well. It is both jargon and a loaded concept, which can be confusing when taken literally and without understanding what it aims to represent.
Funnily enough until they said their country was such, I assumed they were American saying 'global' South to distinguish from meaning a South American country.
'Southern hemisphere' is fine though I think? And I assume that's what was meant.
If they meant it as a replacement for "third world" (in the contemporary sense, not explicitly about the cold war), then they didn't mean anything to do with south as a direction.
It's an outdated term. I recommend reading Hans Rosling's book Factfulness where he addresses a lot of common misconceptions. If anything, it's fun to take the tests and compare where you are against the rest of the world.
Not the person you are asking, but in Eastern Europe we also call our countries "third world"; I learned in the middle school about economic development and classification by that, it was nothing to be ashamed of. Even today being in EU we are considered second class citizens, not a problem with me.
It’s colloquially used to mean economically underdeveloped/developing countries. However recently economists have begun using the term “emerging markets”.
Anything working with (downloading) massive datasets would 'work' slower, but as fast as possible (provided the supplier can keep up too) is better?
Add in a requirement/desire for it being real-time (or equivalently (you'd get further and further behind) continual) and it could be necessary if large (per second) enough.
This. In a developing country, zero internet/mobile to email/mobile SMS has a huge economic impact for farmers who want to check prices in local market. Scale up to 1Mbit via 3/4G, the economic impact is equally enormous for a variety of reasons.
Many interesting things happen at the limit of what’s currently possible. That’s why extreme efficiency of computer hardware and software will always be important.
Also the fact that it would be cheaper for the people who need this tech to just rent out a VPS in a datacenter than to roll out 10gbps to every single person on the extremely rare chance someone needs it.
Telekom is cancer, especially their peering. 250mbits sounds fine though, I haven't upgraded from that yet, there are way too few instances where it restricts me to pay anything more.
Hahaha. And I'm sure the price is cheap. Only an arm and a leg. Oh. And your soul. And with the caveat that the 250 Mbps will only be available in the middle of the night, when everybody is sleeping.
After having had to pay Vodafone 50 euros a month for 10Gb of data through LTE for years, due to the impossibility of getting any connection faster than 3Mbps (even as a business), we are also, through the miracles of governmental financial incentives and magnificent planning (only 2 years late), getting the joys of high speed internet in our area. 100 Mbps. Hoooray!
Telecommunication in Germany is a cancerous, price-gouging cartel that needs to be eradicated at all costs, packaged on the first-available Falcon Heavy flight, and propelled at escape velocity out of our solar system.
Where? I pay 60$ for mildly spotty 120mbps and that was very good deal a year ago. It's pretty sad considering I live in Montréal. I know western Canada has way better internet/data prices though
I'm thrilled with my gigabit fiber from Sonic. I wonder how many of their customers even get the full gigabit though; if they are connecting with WiFi they almost certainly aren't. And even ethernet is still not reliably gigabit in homes with older wiring. But this 25gbit is in another category.
I'm impressed the Ookla speedtest server can deliver 25Gbit.
The 25gbit network card he mentions costs $400-$500, that's cheaper than I would have guessed.
Yeah, I’m a new Sonic customer and got their 10gbit service. $40/mo beats any of the other three providers pricing for much lower bandwidth.
However, I only have a gigabit router, so most of that isn’t utilized, assuming that’s true for all my neighbors as well (Sonic’s only offering here is 10gbit)
Just think of it as future proofing, like installing higher rated cabling in a house than you'll use initially. The cost isn't any more for you to have 10Gbit in this case, so for you it's just knowing that you could buy more equipment for your end at any time to use it if you needed, and not have to wait for a service change. :)
I had Sonic fiber for a few years, in the Bay Area. I always got the full gigabit, basically. Speedtest always showed around 940Mbps up and down. This was over Ethernet, since I always ran Ethernet into the Thunderbolt dock for my main Mac.
Usually got around 500-600Mbps over WiFi on a fast device.
I fondly remember the days in 2010 in Zurich, and cablecom had good functional internet, don't remember the speed. Then I moved to Germany in 2013, and it surprised me that I can't have internet in my apartment for the next two months, because the technician appointment and when I finally received, I got a meagre DSL connection with bad ping and 32 Mbits.
It has gotten better over the years, but even the best available consumer connection is 1000 Mbits down and 200Mbits up.
Majority of the tech Germans I meet are embarassed about the internet and telecom infra, being an advanced economy.
1000/200mbit sounds more than enough to me, even for a family. I could have 1gbps but decided to stay with 250mbit because I just don't see how it would benefit me. Cloud vms are so cheap and tend to have lower ping to other services, so I go there if I really need faster speed.
I do agree about the general state of German internet connectivity though. However, seems to be catching up quickly, especially in rural areas.
The fastest internal SSD in a Mac now can write at 3.3GB/sec, so it would be able to keep up with 25Gbps connections, barely.
Average SSD write speeds are usually around 1GB/sec these days, but most folks could build a RAID of several SSDs and be able to keep up with such a connection, as well.
I would also assume that most folks who would bother to install 25Gbps, or anything close, probably have a lot of devices in their homes to take advantage of it at once.
Personally, I find it very hard to saturate the 1Gbps link I have. Most servers won't push more than around 200Mbps out to you, and I'm basically never doing 5 tasks at once that are that bandwidth-intensive. But it's nice to have some overhead, just for fun.
I’ve had great results sticking a single Optane NVME drive in front of my 8x8GB array of spinning rust via ZFS’s built-in cache facility. I’m only doing write caching and only using 32GiB of my much larger drive (I forget exactly how large lol). That was my comfort zone for how much data I’m willing to lose if something very bad happens to the NAS during a write, and I like giving it a bunch of spare flash space for wear-leveling.
I believe the transfer will simply be bottlenecked, like running a raspberry pi from an sd card.
To fully utilize the connection you can use storage backed by memory. For example ESXi allows you to back a vm by a "virtual pmem" disk that uses a chunk of your host system dram
Hard drives don't go anywhere close, but SSD do. What is important to understand is that 25 Gbps Internet does not mean you copy from a server on Internet to your local SSD at that speed, that 25 Gbps is just the connection between your router and the ISP. On your side you may have multiple computers, on the ISP there are multiple peerings with other networks, you may reach and aggregate bandwidth of 25 Gbps but not really a point to point one, so your SSD performance is not the top factor.
25gb/s is nice. I am envious of locations that have a modern local fiber network and more options.
My 2 fiber connections are 500mb/s and each cost $350 to activate, $150/mo each not counting the costs for static IP's. Trenching it in was $3k. I'm more than happy with 500mb but the price could be lower. For my specific use cases it is less about bandwidth and more about latency/peering arrangements. I would personally be happy splitting as low as 100mb/s across many devices with fq_codel or cake and cdg on my little firewalls. I can't really complain though. The alternatives here are 4G LTE or Starlink.
What is the use case for this? How are you doing to utilise it? According to fast.com I have 29Mbps and I see no reason why I would need to go faster now. And to be said, I'm WFH all the time, I have a homelab and we stream movies, not once since pandemic started I had a thought I might need faster broadband. My home router has SFP port.
At 25 Gbit/s your computer may likely not be able to keep up, unless you have NVME based SSD. We are talking about 3GB/s+ which my 24-drive NAS (old) can’t get beyond 2.6 GB/s with ZFS.
Absolute madness. But this kind of bandwidth isn’t meant for a single machine.
In the end it's just an arms race with basically no use cases. You could call it future proofing but the switches will have to be replaced before such speeds could realistically be used (8k/16k video or sth like that?)
Having fiber without pon is great, means no new infrastructure for decades. But anything above 10gbits will be too much to use realistically, even for smaller companies.
Well, the use case is multiple people doing things at the same location that add up to more than 1Gbit. For example, think of the connection on a switch with mostly 1Gbit ports but one or two 10Gbit or 25Gbit ports for the uplink. Switches such as this will also have a backplane capable of doing more than 1Gbit.
Individually no one person/port can use more than a gigabit and can't saturate the switch in this case, but combined they could utilize far more.
25Gb/s is just overkill for residential use. It's really cool that's it's available, but I fail to see a use case over my 500Mb/s home connection. Even for the servers that I manage and that are bandwidth-heavy, 10Gb/s is way overprovisioned for now.
WiFi goes up to 1Gb/s, if you're lucky. Sure, some WiFi-6 APs have a 2.5Gb/s connector, but that's not what you want or need, unless you're a high-density enterprise. WiFi-6E will possibly improve that a bit, but it will take WiFi-8 to get anywhere close to saturation.
Wired, you can do 10Gb/s for server systems, which are, amongst other things very loud and not very suitable for placement anywhere near humans. 2.5Gb/s support is spotty, and 1Gb/s still the only thing that works reliably.
So, exactly which residential application requires 25Gb/s is not very clear. Yes, it's cool, but not very useful, and faulting manufacturers (especially in times of crippling supply-chain limitations) for not fully supporting it is questionable.
Without drowning in fan noise? This world... Sure, I guess you can get a Mac Studio, or some other 'workstation' class PC, but your switch will still need to be within a few meters of that endpoint, and it's not going to be very green nor silent.
2.5Gb/s can easily be done with a lot of laptops and workstations these days; 10 Gb/s isn't quite there yet (and 25, 40 and 100Gb/s are definitely in the server-only fiber-or-DAC-only realm)
Mikrotik makes a whole bunch of 10G hardware. The 8 port ones are fanless and have a heatsink on the back, and the 16 port model has a fan that only gets switched on when needed. I have one sitting behind me and it's been off all day.
It's also easy to open and replace the fan with something less noisy. Just remove a few screws, it's a standard size with a normal connector on it.
There's also big external heatsink so you could rig up a big, slow, fan to help it out a bit.
If you want to be green by the way, use fiber/DAC. The 10G copper SFP+ modules are power hungry, and Mikrotik recommends not placing them next to each other. Also the extra power draw is likely to result in fan use if you have a lot of them.
I had issues with 10G copper SFP+ modules even when there was only one installed in the switch (Mikrotik CRS305-1G-4S+IN), the other modules installed being DAC/fiber. I got random disconnects I could not attribute until I checked the switch logs - the module was shutting down because it was reaching >90C when the ambient temp was 26C. I had to add a fan.
Not all transceivers are created equal; some need more power (and then dissipate the heat), some are satisfied with less. Some can do only 30m distance, others will run over 80m distance.
The Mikrotik ones (S+RJ10) are based on Marvel chip and they are the more power hungry / run over 30m only variety. On the other hand, they can negotiate 2.5G or 5G if necessary and support a proprietary protocol to tell the switch about it, so you will that in SFP+ properties.
As I have written in the sibling comment, I have good experience with BCM84891-based transceivers. CRS305 can handle two (still not next to each other, obviously).
Oh awesome, and pretty well priced! Know what the noise level is like? The QNAP is designed to be in a home office rather than a server closet, so it’s inaudible. But my mikrotiks are passively cooled, so even better.
> but your switch will still need to be within a few meters of that endpoint, and it's not going to be very green nor silent.
Mikrotik CRS305 (4xSFP+) and CRS309 (8xSFP+) are both passively cooled. They are pretty much silent :) though the blue led takes some tape to be less shiny.
> 10 Gb/s isn't quite there yet
If you really, really need RJ45, look into BCM84891-based transceivers. They still get hot, but not as much as others (according to datasheet, takes 1.6W at 30m and 2W at 80m). I also managed to get stable 10g over 20m Cat5e with them.
There are a few quiet 10 GbE switches, the passive Mikrotik ones others are mentioning, but this one is a quiet actively cooled one if you want RJ45: https://www.qnap.com/en-us/product/qsw-1208-8c (Lots of combo RJ45/SFP+ ports)
1 Gbit/s is pretty close to still being overkill. 25 Gbit/s is laughable overkill and appears very likely to remain that way for the coming decade.
Consumers around the globe have had increasingly common access to 1 Gbit/s for a decade and there still aren't any other great, common use cases for it beyond very high quality video streaming.
It didn't take very long for computer use to need more than 640K by comparison. In the computer realm those edges were being constantly pushed at that time. Such is not the case with bleeding edge consumer broadband speeds today.
1 Gbps is slow. Even recent wifi can plausibly exceed 1 gpbs to a client. Pretty much any modern HDD (let alone ssd) can read or write faster. USB 3.1 is an order of magnitude faster and display port 2.0 tops out at about an order of magnitude faster than that. It doesn't really make sense to leave the main physical network interface so far in the dust, let alone claim it's overkill.
1 Gbps is laughably slow. It doesn't even keep up with hard disks. Network attached storage is crippled by 1 Gbps networking. It's ancient. I remember doing an assignment in 2005 to design a network on a budget, deciding to "splurge" on gigabit, and finding it very much affordable. That was 17 years ago, and yet consumer networking barely budged since then.
10 Gbps is still below the 7 GB/s that a single NVMe on PCIe 4 (which is readily available) can achieve.
25 Gbps is still below that.
I'd say 100 Gbps is where the current practical maximum is more or less. You'd have a hard time writing or uploading data that fast on anything resembling consumer hardware.
25Gb NICs are getting into the affordable range, but SFP28 modules are still very pricey in comparison. I’m a big fan of the Mellanox CX4 LX cards for their low power draw (11w max). OEM cards come cheap and can be crossflashed to generic Mellanox firmware.
The bigger issue with 25G is that it’s well into the range where any OS’s default TCP settings won’t provide anywhere near line speed, and once you solve that it can expose other non-network bottlenecks. I have dual-SFP28 cards in both my workstation and NAS, both connected to the network via a10G Mikrotik switch and directly to each other via a 25G point-to-point link. Now after all that tuning I have a fast network but run into the SAS bus bottleneck for any file transfers exceeding the size of my app take write cache :p
>but I fail to see a use case over my 500Mb/s home connection.
The last time I thought I have not enough BW on my 50Mbps connection I learned what it's just my T440 is not enough to decode h265 extremely compressed stream over an SFTP streaming over WiFi, which gave around 20 Mbps at best. After toying|fighting around with wireless settings I found the main culprit was my Intel 7260 (or whatever) WNIC, not the CPU, Internet connection BW or the server throughput (10Gbit, despite being an IIS instance).
521 comments
[ 2.8 ms ] story [ 337 ms ] threadIt's most likely an operational mistake - have you called and asked for them to nuke the line and then set it up again?
Or are you saying that you are getting your advertised speeds, and those just didn't go up when you "got fiber"?
Don't know which one is more worrying if true, to be honest.
I think beyond 1gbit, the benefit become super marginal and the hardware expensive.
Probably the best argument would be that many/most? people don't watch the whole video.
The main argument I see is "if you build it they will come" i.e. we won't see higher bitrates until connections like these are more common, but for now gigabit has even fairly extreme use cases for video streaming pretty well covered.
Past even 500 Mbit I'm way more interested in latency considerations than raw bandwidth, and practical matters like how to use that bandwidth from my laptop (good luck doing 500 mbit wireless reliably, never mind 25 Gbit!)
I'd still rather have gigabit fibre.
The fastest I ever saw on 4G was 300 megabit down but the latency could be 40-80ms.
Most routers and wifi adapters are crap. Buyers do rarely go beyond "wifi 5" or "wifi 6", and do not realize that there's much more.
The older Apple Macbook Pros (pre-2019) came with 3x3 MIMO ac adapters. If you had capable AP on the other side, you could reliably do gigabit with them. The newer ones have only 2x2 MIMO, just like the rest of the laptop market, so you will get only 600-700 Mbps (out of theoretical 866 Mbps).
If you are getting 500 Mbps and there's not a concrete wall between your client and the AP, something is quite wrong. Misconfigured AP, your client cannot do multiple streams (yes, there were adapters like that sold on the market), or just older/pre-ac AP or client.
Would you mind elaborating on this? I would like to learn more to revamp my home setup, which is currently on wifi 5.
1) how effectively you can pack data into a channel, determined by the modulation scheme.
In your wifi properties, you can see the modulation scheme used as 'MCS index'. You will see an integer between 1 and 11, the higher, the better. This is negotiated between client and AP, depending on the signal strength (antennas, number of walls/wall material between them, etc). With wifi 5, it is realistic to have MCS 8 or 9, if your client and AP are in the same room.
2) how wide is your channel - the "basic" is 20 MHz wide, newer standards can combine multiples, into 40, 80, 160 MHz wide one. Wifi 7 will bring 320 MHz wide channels.
2,4 GHz band has only 3 non-overlapping 20 MHz wide channels available. If you use more than one, your neighbors will hate you. The situation is better with 5 GHz band, unless you run into the DFS (radar detection), then you can have intermittent outages. This is improved by Wifi 6e which brings more channels in the 6 GHz band, but the client and AP support is not there yet. Currently, running an 80 MHz wide channel bellow DFS frequencies is probably the best.
3) how many many simultaneous streams you are able to transmit/receive - MIMO (one client using multiple streams), MU-MIMO (multiple users using multiple streams, so they do not stomp on each other and sidestep next point).
Most clients are capable of 2x2 MIMO, older Apple MBPs (pre-2019, -ac based) are capable of 3x3. Most APs are also 2x2. The nicer APs can do 4x4 MIMO, some gamer APs by companies like ASUS can do 8x8.
4) how effectively can multiple clients share the same channel.
Here the problem is, that once an older client connects, it is downgrade for all clients, they must be compatible. If you can, keep your 2,4 GHz network -n and higher, and 5 GHz -ac or higher.
Once you will be able to have a band -ax only, you could theoretically be able to use OFDMA, which does improve the performance for multiple clients.
You can have a look at mcsindex.com: pick the modulation scheme (row), number of MIMO streams (group of rows), how wide is your channel (column) and OFDM/OFDMA (group of columns) and there you can see the theoretical bandwidth. Note, that it is a shared bandwidth, for all clients.
There are some minor points:
5) if you have IPTV or another application that uses multicast, do not run it over wifi! Multicast by definition does not have ACK, so what can AP do to ensure every subscribed client gets the data? Slow everything down, for everyone. So do not let it do it.
6) multiple SSID - possible, but do not overdo it; they also have performance impact (the SSID advertising takes time, at the slowest/most compatible rate allowed; the more of them, the more time it takes). Ubiqiti limits them to 4 per radio.
7) at higher frequencies, you won't have a such range than at lower ones. This is both good and bad thing. The bad, obviously, that your range is lower. The good is, that the range of your neighbors is also lower, so you could use the same channel without disturbing each other.
8) It does not help to crank the tx power to the max. The clients may receive, but when they answer, the answer could never arrive. Clients do not have antennas or tx power like APs do. Use the tx power reasonably; if you have multiple APs in your house, adjust it so clients will let go the weaker signal and connect to the stronger one (which AP connect to is fully managed by clients, unfortunately, and they often prefer to stick to the one they are already connected, even if different one with stronger signal is available. At most, some APs can kick the client once the signal is below threshold you specify).
9) some cheap dual-band APs have only single radio, capable of transmitting at both frequencies. Not at the same time, obviously, they time share. Make sure your APs have separate radios for each band they support.
10) some APs do support roaming (802.11-k/r/v); for home (or WPAx-Personal) that...
Is it expensive though? We spend how much of our lives online. How much do you think this all-spanning life upgrade costs? What is the price of never ever dealing with buffer-bloat? Take a guess.
I calculate it as a one time $356 cost plus labor. You might have paid more for your wifi system. 128 port 25gbe switches are around $20k ($156/port). Transcievers are under $100 and you need one on both ends. For a lot of already deployed fiber, this is a drop in replacement. This is absurdly cheap. Given how cheap this is it's an obvious & enthusiastic heck yes. Who wouldnt throw down $356 right now to get 25Gbe for life?
Cost gets a bit more complicated when it comes to the POPs & their uplink. Subscribers are going to be way oversubscribed even with some fairly expensive 100Gb uplinks. As you get further from an exchange the difficulty grows geometrically (because pops become.further hops away from the ix). Peering needs to be bigger too, as does transit (but ask whether the net volume of traffic grows elastically or not), which has costs. But I think we need to frame this question a bit better, of whether it's "worth" upgrading. Honestly costs are so low it doesnt make sense not to; the rest of the world is just milking us, bilking us, charging what the market will bear, protecting it's profit centers, and this company init7 is doing what makes financial sense for the consumer. Donwe need all that? Maybe maybe not. Should we settle for less? There's almost no financial case when the hardware is so so so very cheap. This tech sounds magical but 25Gbe is not exotic, not extreme technology; "the future is already here, it's just not evenly distributed yet".
To actually realize your faster speeds, you need to spend thousands of dollars yourself on new switches and NICs. And then, as mentioned, the benefits are marginal. You would have to be streaming 10+ 4K movies at once to even "need" gigabit, let alone 25Gbps.
I semi agree that I dont think we know what this is for. Never ever having buffer bloat is a tempting first ask. Connectivity is more than the sum of throughputs, as your figures imply- there's questions of availability too.
Being able to access each other's systems at near local speeds sounds quite compelling, could help jumpstart post-Big Social computing. You talk about netflix streams, but those are heavily compressed with the best offline encoding on the planet: if i just want to open Steam Remote Play Together & share realtime 4K with a friend, I'd need a lot more throughput since I have much much much less efficiemt encoding. If i wanted Remote Play Together with 3 friends, well, that figures goes up. If my family member also wants to do the same, now we're using a lot or maybe all the throughput & we're starting to have some contested bandwidth, some rising latencies.
The truth is somewhere between. Rationalizing ourselves down to what sounds sensible today, to me, is a cruel trick, is not just path dependency but an ideology that believes only in what we have & can see now, & refuses exploration & trying. To me the world & tech is spiritually fueled by why not thinking, by deciding to opt for the extra thats within reach.
Forgoing a cheap (still less than the price of a nice tv, by far), available one-time purchase option that vaults us into near-local connevtivity caliber with the world is still a lock in my book.
There's benefits to 25G (certainly when transporting 4K video around which needs more than a 10G nic), whether that's worthwhile for a typical home is likely "no", so unless you've got hundereds of employees in an office it doesn't feel very useful.
Don't get me wrong. I'm very much in favor of this upgrade. I just don't think it's going to be an everyday quality-of-life improvement for most people. It's more about providing a service for people with special needs, and future-proofing the infrastructure.
Indeed. And I'm speaking as someone who downloads most visual media before I watch it, so bandwidth matters to me. But not that much. With 100-200 Mbit/s I am good.
Cookie prompts, newsletter pop-ups, scrolljacking and ads constitutes the majority of wasted time for me, by a long shot. Latency to sites in other parts of the world can cause problems sometimes, since number of round trips can be quite high with TLS neg + progressively loaded content.
Unfortunately for most (consumer) FTTH deployments, that's not the case. Most of them are GPON, where the initial deployment was more effective, as up to 64 of your customers share single fiber, but then that means all of them have to upgrade all 64 of them at the same time, you cannot do them one at the time (see also the speed of XG-PON upgrades).
Additionally, many providers forced use of their CPEs. If you can send out SFP module and the customers can put it into whatever they want, it is much simpler, as replacing CPEs for all the customers on that fiber.
As TFA mentions, same was the case with 1 gbit. I can confirm, I'm on the same ISP and was "early" to have 1gbit/s. Lots of servers still only had 100mbit/s links. These days I have zero issues saturating my 1gbit link.
Give it time, now that it starts rolling out, costs will come down and server links will be upgraded. The usual early adopter stuff.
http://tm.kit.edu/doc/2019-LCN-100g-tuning-authors-copy.pdf
Some time ago PC has revolutionized the world by giving access to computer power to general population. This has unleashed a tidal wave of creativity and business.
Giving the ability to host own servers to everyone can open up endless opportunities as well.
I host some of my own servers and benefit from it greatly.
It’s my biggest hope for internet re-decentralization.
Software and driver updates exceed 1Gbps all the time, as do game updates/downloads through Steam.
Piracy also works really well. Downloading copyrighted media is perfectly legal in Switzerland and I was able to get ~7Gbps real-world speeds from Usenet without too much hassle.
It's also really handy for things like backups. As of writing, bandwidth to a Hetzner cloud server is ~5Gbps up/down with iPerf3.
I do agree that 25Gbps is more overkill/bragging rights than it is real utility but I think 10Gbps is an easy sell.
Keep in mind that there is no change at all to the monthly price of your internet by choosing these higher speeds. 10Gbps has the same monthly and setup costs as 1Gbps, you only have to pay a bit extra for hardware capable of dealing with 10Gbps, which is pretty affordable.
Please note too, piracy is not legal in Switzerland. What is illegal is the spying and tracking which would be necessary to build a case and prosecute pirates. Technically you could still self-incriminate if you documented all of your piracy, with proof, and published it online.
The mentioned legal basis is [1].
Note that this applies strictly to downloading though. Participating in a torrent swarm (where uploading is also happening) is not permitted. That's where the technicality you mention comes in: until recently, it was illegal to monitor internet users for copyright enforcement purposes, which meant it was illegal to monitor a torrent swarm, which meant you could somewhat safely seed torrents, despite it being illegal.
That loophole was recently removed however.
[0]: https://www.ige.ch/en/intellectual-property/counterfeiting-a...
[1]: https://www.fedlex.admin.ch/eli/cc/1993/1798_1798_1798/en#ar...
De jure it's not really legal anymore since they updated the law... De facto nobody cares as long as you're not making a business out of it.
Same goes for the place I live since 2016: Thailand.
Though: Not pirating for professional stuff, only for private stuff. I still support software/media by buying the things I want to support.
My claim was founded on authoritative sources, including the law itself. If you want to contradict that claim, you'll need to back it up.
The realistic limits are per connection. But with 25Gbps you can just have multiple connections open without any of them ever affecting any other.
Realistically, this is pretty close to true with 2gbps symmetric, too. My provider seems to give me 2.2gbps in practice.
PS5 downloads/updates are 500-600mbps in practice. Steam is 1.5gbps or so. Most other things-- streaming, video calling, etc, are under 40mbps. So, you know-- if I kick off a PS5 download, and a steam update, and my kids are streaming and video calling... And my machines are backing up to the cloud at 1.5gbps (other direction)... and I decide to do a big apt-get update on a machine, maybe my steam update completes a couple seconds later.
Of course, I want even faster... but I'm hard pressed to say what would be better.
SATA drives max at 6 Gb/s, and probably a bit worse than that in practice.
M.2 NVMe SSDs can reach 25+ Gb/s on sequential writes. Maybe I could overclock these to squeeze a bit more out too. Although, an 8TB M.2 drive will cost over $1k for now.
I suppose there's also RAID. Although, I'm not sure what other limits/gotchas I'd run into there.
Quite frankly, I'd expected a 100Gbit uplink by now.
I just got a notice in the mail that my ISP is "upgrading" their network so now I can pay $200 USD/month to get a whopping 2gbps, which I actually thought was pretty amazing until I read your post. So, thanks.
In all seriousness, congrats! You made a good case for why one would need that much bandwidth. Also, we need to catch up here. :)
Until recently I paid $35 per month for a 300 Mbps line because I couldn’t justify the jump to 1 Gbps.
Then I had to upload some local Hyper-V servers to AWS to convert them to AMIs and figured what the heck I’d upgrade the line.
I would have no interest in upgrading beyond 1 Gbps right now though. There are too many infrastructure components that need to be upgraded to attain that and I don’t have a use case for it.
My local ISP trenched fiber to my house for free and provides gig internet for $80/month.
The funny thing is my previous house was in town and I had to settle for 100Mbit for the same price. ISPs are all sorts of messed up.
In the city it’s often just as easy to let what is working continue “working” - a major rollout takes a lot of money.
(I'm on Comcast's Business service, $250/mo for 1000/35 [long dumb story why]. Most of the time I see under 600 down when checking on speed test sites, and real-world speeds downloading large files rarely exceeds 250. I expect real-world speeds on the non-business service are even worse.)
It's pretty embarrassing that this is the state of things.
Over in Oakland I am paying $40/mo to Sonic for 10Gbps (though I only have equipment to route at 1Gbps at the moment)
There's a benchmark posted there showing that the speed is really as advertised.
gnpUSD = {'USA': 21,650,000,000,000, 'THA': 491,910,000,000} #4th quarter 2021
pop = {'USA': 329,000,000, 'THA': 70,000,000} #as of 2021
gnp['USA'] // pop['USA'] 65805
gnp['THA'] // pop['THA'] 7027
Seems more like a distribution of resources issue... so why not jack up taxes on the wealthiest 1% and use it to pay for things like high-speed fiber in all the rural areas? Doubtless this would lead to economic growth?
Throwing more money at it usually makes that kind of problem WORSE not better.
> "The Rural Electrification Act of 1936, enacted on May 20, 1936, provided federal loans for the installation of electrical distribution systems to serve isolated rural areas of the United States. The funding was channeled through cooperative electric power companies, hundreds of which still exist today. (wiki)"
However there are just as many places where the state's government was bought off to ban such networks because the majors are afraid of actual competition.
It’s mostly been ineffective.
---
footnote here: official income record of thai people/businesses does not really reflect reality. Lots of entities are kind of off-system.
I prefer to order grab of around 120-200 baht from a "healthy" place or going for restaurants within Central 200-400 baht per meal.
It's like a typical consumer in the US spending $500+ per month for broadband. Even worse if you consider the rural income factor. Absolutely insane.
For a fraction of that you can get 1gbps from Comcast and you'll never utilize most of it in 99% of consumer situations.
I'm in a small quasi rural 'city' in the US, hours away from any consequential city, and I can get 1.2gbps from Comcast for 15-20% of the income adjusted rate in question referenced for Thailand.
It seems common to forget how astoundingly high US median income, disposable income, and GDP per capita figures are compared to the rest of the world. The latest 2022 estimates are pegging US GDP per capita at nearly double that of France and Japan. To match up on median disposable income figures, you have to use hyper affluent countries like Australia, Denmark, Sweden, Switzerland and Norway as references. Then people come on HN and proclaim how they're paying only $20 per month for Internet access, in a country with 1/10-1/5 the median disposable income of the US. The US is more expensive than it should be for Internet access (better telecom competition would go a long ways toward fixing that), however the reality is US income figures are also a lot higher than most of the developed world.
Where your number comes into play is for people pricing Netflix subscriptions.
You end up with absurd examples where Botswana is comparable to China; Russia is comparable to Greece; Puerto Rico is comparable to Spain, ahead of Portugal, and just a bit lower than Japan; Kazakhstan is just a bit behind Latvia and Slovakia; Taiwan is far ahead of Finland, France, UK, New Zealand.
Imputed rent for example is one of those things that’s kind of silly on the face of it but makes various comparisons more reasonable. On the other hand it can also imply a great deal of economic activity that isn’t actually happening.
PPP is the same sort of calculation. If rents crash because a great deal of housing was built it can make GDP comparisons kind of meaningless. The country has more tangible wealth, people are better off, yet GDP falls. That’s not what you want the number to represent.
That means you can have two countries with comparable GDP per-capita, where one of them have a more affluent population and able to pay higher prices.
I went with 5G from T-Mobile for $50 a month.
I live in rural BC and I'm was able to get 80Mb for $35 using a no-name internet only provider with no contract.
If I went with one of the big cable companies I would be paying at least double for the same thing
So my internet is provided by Lightspeed on Shaw cable.
As much as Canadian telecom seems to own the CRTC it isn’t nearly as consumer unfriendly as when I lived in the US.
I had an apartment in downtown Seattle which had 2 choices for internet. Many of my friends were quite impressed since most addresses are only served by one ISP.
This price randomness never occured to me in germany, and just booking a fixed low price on a website was so much more convenient.
Moved to Thailand (BKK) in 2016 and was so disappointed by a very new condominium in the center, which didn't even allowed for FTTH lines.
Real Estate Developers usually make contracts within their Condos, to force the whole building to use only one provider.
But it doesn’t matter because it isn’t actually offered in my area / the local telco can’t even notify the right people…
Also amusingly, both Comcast and Verizon have the very old (as in multiple decades) name for the street I live on as my address. It's only somewhat wrong (the name used to have a North on it)--but it is still wrong.
It's orders of magnitude more uncommon than doing so in the USA.
So bring all the hate :)
Where I’m at it’s $80/month for a 5gb/s. If you want that in a major city, >1m people; good luck.
Corruption, regulation and development costs are just too high.
Rather sad to be honest.
Mobile plans are also on the cheap side, to the point of being competitive with many third world countries (for example, I wasn't that impressed with prices in Thailand in 2020 for comparable plans, only a 1-2€ difference with what I had at home).
This is getting to the point that I find myself often suggesting fellows from border countries open a line in France, just so that they can enjoy the "European roaming data envelope" that comes with most plans (i.e. several gigabytes are free to use from anywhere in Europe, and you can make calls and texts to other EU countries when abroad...) after realizing how expensive and suckish mobile plans are in their country (Belgium in this instance).
[0] https://www.connexionfrance.com/article/French-news/Best-and...
Which provider would you recommend? I remember taking a look a while back and the most viable option was Free, but 15.99 EUR a month for when I need roaming 3–4 times a year seemed a bit excessive at 180.- per year.
SIM card is free, you can top it up with a prepaid amount or do a one-month plan without renewing it, and I just checked their T&Cs relating to roaming and they are actually quite generous[1][2].
Besides, their website is (for the most part) translated in English due to their target demographic, so no need for full French fluency here!
Some good info in English on the French telco landscape can be found at [3].
[0] https://mobile.lebara.com/fr/en/
[1] https://mobile.lebara.com/fr/en/international-call-rates
[2] https://mobile.lebara.com/fr/en/terms-and-conditions#section...: See section 2.4.2 (in French unfortunately)
[3] https://en.selectra.info/broadband-phone-france/guides/mobil...
Edit: missing word
A cursory search indicates that both Orange and Bouygues (as well as their contract-free offshoots, Sosh and B&YOU respectively) still offer this as a paid option at 5€ and 7€ per month respectively.
This means you could get the plan appropriate for your needs, then suspend it (for at most 6 month per year it seems) and only reactivate it when you need it.
However since this procedure requires getting in touch with customer service, it should be reserved for people well versed in French :-)
Edit: adding links (all in French)
https://www.echosdunet.net/dossiers/suspendre-ligne#suspensi...
https://assistance.sosh.fr/faq/308268
https://www.assistance.bouyguestelecom.fr/s/article/suspensi...
If you know someone in France, you may ask them to make a "solemn declaration" that you live at their address; then you only need provide their own proof of residency (people living with their parents routinely do so).
Tourist and prepaid SIM cards are still exempted and, depending on your case, can prove quite interesting.
For example, we offer $40 service, and if you're in our historical areas that's 1Gbit symmetrical, and if you're in newer areas we've turned up in the last year that's 10Gbit symmetrical.
We're expanding (as I expect most ISPs that can undercut the major players that much are), but it requires actually stringing fiber through neighborhoods on poles, so it takes a while (but that can be scaled...).
If you don't have telephone poles, it's much harder/more expensive to build out an area, so often those are skipped (at least initially) as areas cheaper to deliver to are prioritized. The is unfortunately a lot of new development, as they'll build neighborhood with underground utilities and pre-wire AT&T and Comcast, making it hard for others to deliver to the area without a lot of cost and work (trenching).
Some people a few miles away have a fiber POP at the end of their shared driveway, and are trying to decide if they should pay the extra couple hundred per house to go from 1GBit to 25GBit symmetric to the houses.
At the ISP I work for, I believe we spec out the cabling we need to a neighborhood and then order a special bundle with breakouts at specific locations along the length to serve locations, and then string that along the poles. I'm not sure why we wouldn't serve a house with one of those, but those go back to a central point in a neighborhood, and it's possible the backhaul from the central office to that central point in the neighborhood passes houses that aren't served. Where to build is all about ease of wiring an area and housi g density. It's all about cost per houses passed. The good news is that maybe your area is slated to get fiber since it goes by there, and it's just a matter of the lower hanging fruit being picked first.
This street got sold off to a bankrupt telco (from what I can tell, their business model is to buy up unprofitable lines, then periodically file for bankruptcy).
There are a few local fiber providers. Maybe I could call and see what it would take to convince them to build out. (There is definitely demand here.)
Alternatively AT&T (who is not our local telco) keeps spamming us to get a business fiber to the home connection. They make it clear they'll ban customers that appear to be using it for residential traffic, but I think that sort of traffic discrimination falls afoul of California's network neutrality laws. Has anyone tried calling their bluff on that?
Be careful approaching incumbent major ISPs or letting them know if you got the signatures though. You might find they went out and offered sweetheart deals to the same people on two year contracts, but at existing service levels...
Regarding business fiber, I think they're more trying to keep people from reselling it as a micro ISP. If you're willing to pay business costs for personal use, I don't think they'll care. If you split the cost between 10 neighbors, they might. There may be other concerns they are dealing with that I'm not thinking of though.
For example I have 3 Internet connections at home that I pay ~ $30 in total, from 150 Mbps to 1 Gbps. I think the 1 Gbps is ~ $12/month, it is so cheap because there are so many options and I can afford to keep all 3 for redundancy (1 is a 4G mobile data capped at 200GB/month, it works even when there is a power failure in the entire neighborhood).
But if there is no variety of options and no competition, then price is high. From what I read, most ISP in USA are squeezing as much as they can from their clients, sometimes to ridiculous levels.
Here in Australia: $90/month nets you “100/40mb”, which is really more like “~85mb down, 25 up but only a quiet night, if you’re next to the router, and it hasn’t rained within the last week”.
If you live in an area that can get HFC, FTTP, FTTN etc. it’s possible these days to get a pretty consistent 1000/50 for $150/month. Those fibre services are pretty stable in my experience, definitely a million times better than the old broadband which as you put got destroyed by bad weather and distance from the exchange.
Apparently you can pay money to have fibre routed to your house from the nearest node/etc, but I rent, so I’m not going to do that.
I'm currently on 1000/50 cable internet, which is already quite nice. Telekom is laying fibre but it's not clear yet whether or not they will stop short of this house. Also i suspect it will be a while before they offer better than 1000/500 service.
https://en.wikipedia.org/wiki/DOCSIS#Throughput
Eg, your tech and cable can support 40 Mbps. If you split that 20/20, your users will have trouble watching 4K video that wants 25 Mbps. Change that to 35/5, and for most people it'll actually work better.
This issue goes away eventually as you either get a fiber for each direction, or you can just push terabits through fiber anyway, so there's plenty capacity to be symmetric without any compromises.
You have one cable, a given frequency may go in one direction or in another. Both sides have to agree on what it's being used for. You'd need a communication channel between the ISP side and the client side to constantly negotiate, and that negotiation would take some time, so such a mechanism would have some latency to it, with possibly weird effects on things like online games. You could get weird behaviors where some particular pattern of traffic would result in the connection readjusting itself just wrong on a regular basis and result in hordes of angry gamers.
I think it's reasonable to guess that ISPs targeting consumers have no interest in monitoring and troubleshooting such a thing when they can just set a fixed split and be done with it (and ask the customer to upgrade to a bigger plan if it's not good enough), and ISPs targeting enterprises have no need for it.
Edit: And there's the issue of how you sell such a thing. It's a system that readjusts itself automatically based on some arcane magic and may work differently from one day to another. How do you make any reliable promises about it?
But I'm spoiled, growing up the only ISP in town spoonfed everyone higher and higher speeds even though no one asked for it. They're laying fiber to villages seemingly just 'cause. People are choosing 4G over fiber because it's cheaper (even though data is limited), go figure.
Well, actually there was a bit of government initiative (with no funding or enforcement) to boost the IT sector. The US would benefit much more heh.
Perhaps it's a conspiracy to have less tech-savvy work-from-homer's keep bumping up their speed plan because their upload speed isn't up to snuff for video meetings. Either that or cable companies feel they need to compete on advertised download speeds with fiber, which is a losing battle.
65 CHF = US $67.89
If you live outside the city you may not be so lucky and spend 100+ on maybe 500mbit internet.
Init7 is also currently in a legal fight with swisscom because swisscom wants (has already started) to pull some alternative fiber that doesn't directly connect the customer to the pop therefore preventing competitors offering faster service than swisscom. The small cost savings that swisscom has is a huge problem for future upgrades and requires power in underground shafts. Not very green when you could pull the fibers directly.
When we got 2Mbit I was AMAZED. The entire internet now actually semi-worked.
Now in 2022 I have gigabit (down, 50Mbit up) internet, with WiFi6 on my devices it's amazing (I'm hardwired for my desktop).
Aussie Broadband here in Australia is my current ISP, and they are amazing. FTTP was a major upgrade.
But the thing which hurts where I am (Perth, AU) is the latency to everything, not the download speeds :).
You have 100Mbit internet? Cool, your 54Mbit only WLAN devices are now the bottleneck.
But now it's your 10GB ethernet that's the limit.
I just upgraded to 10GB networking in my house and the old desktop I have hanging around mostly for guest use tops out at around 6 Gbps. The CPU just can't handle more than that. Granted, it's a 10 year old, cheap CPU with integrated video, so it was never much good. But still, it works just fine for web browsing.
And of course hard disks get 200MB/s on the very rare occasion, and often a good amount less than that. Even SSDs are limited by SATA's maximum of 6 Gbps, so make a plan for upgrading everything to NVMEs.
Curious thing, the ISPs don't oversell even though they easily could. I get exactly what I pay for, speed is never below ~940Mbps (down or up), and uptime has been stellar.
But, of course, I'd still love 25 gbps fiber to the home just because.
Back when covid19 started, we all went home, the office closed, and then we had a server failure. And I spent multiple days downloading stuff to back it up before reinstalling the machine, because there was not enough spare storage inside the office.
If you work with VM disk images even 1 Gbps starts feeling a tad sluggish.
I work with VMs. Disk images stay in the datacenter, and occasionally get sent between datacenters. They don't live at my house.
What I do enjoy is that rsync.net is using the same ISP so I can max out my upload to them.
My speeds are advertised as 1Gbits down and 100Mbits up. On speed tests, I get much higher downloads (~1.5Gbits per second). This might be because it's not capped properly, and I live in a place that they're still building houses in (I imagine that our "box" is not yet saturated).
However, I rarely max that speed out in any download. Video games on Steam max out at maybe 65 megabytes per second. wget commands go from between 10 megabytes per second to maybe 40 max.
The thing is, I know that the speed tests are legit, because I can do several of these things at once and none of them lose any speed.
Copper or radio connections just don’t seem like as worthwhile an investment to me. I’m not saying it makes sense to have an 800gbit connection to a residential property but from a longevity, cost per mile, use of less rare materials, reduced rf interference or whatever, fibre just seems a better proposition to me.
That, along with a surplus server I literally housed in my shoe closet, gave me the firepower I needed to prototype out something that led to two research grants which now employ myself and a new PhD student.
We discount technological investments like this as being “too much” and “unpractical”, but we forget that, even if it only one person in a hundred or a thousand take advantage of them, the impact can be enough to launch careers or start businesses with sizeable positive externalities.
But I really wonder what was your prototype that needed that speed connection and that wouldn't work with something slower. Can you share, at least vaguely, what was your work?
We work with gis/satellite data. Nothing groundbreaking technically, but we aggregate it into products that are useful for social science researchers who aren’t comfortable with or don’t have the computing resources to use this data by themselves.
We’ve actually moved everything into a data center with a 1gb connection. The trade of being that we have several orders of magnitude more storage and computing capacity.
I was surprised when my parents had 1Gbit in 2009. I was delighted when I could get 500Mbit in 2013. I'm slightly miffed that I don't even get 1Gbit in 2022. I literally laugh at providers attempting to convince me to get whatever with 10 or 20Mbit uplink speeds, in 2022l Yes, they exist.
Anyhow. What's insane about a 10Gbit uplink? I wired up my apartment for 1Gbit in 2001. I've been frustrated about home network speeds for 21 years. I do not understand why you consider 10Gbit insane.
On the wider internet, few resources are happy to dedicate such bandwidth to you, even if you can consume it. Not even every popular torrent can muster so much incoming bandwidth.
I can see how it may make sense if you run a popular public server from your home; maybe more convenient than colocation, but still is not a typical use case.
Real question: If you are an Internet gamer, I assume that latency per packet is lower for 1Gbit+ connections, so it is useful for many customers.
More: I feel the same about monitor resolutions above 4K. Anything above 4K on a 27inch monitor makes no sense for my body size / sitting height. I cannot use a vertical monitor taller than 27inch, and above 4K is too small for my aging eyesight! Again: I am sure there are many interesting applications for 4K+ and 27inch+. I hereby wish good luck to all the whippersnappers who will prove me wrong with interesting applications of 4K+ and 27inch+!
So? The process of wringing out the bottlenecks is to eliminate the ones you can, and be mad about the ones you can't.
Back in the dialup days, nobody in their right mind would run a torrent, or play online shooters, or watch streaming HD video -- such things were utterly impractical over ultra slow links with very high latency. Once people started getting broadband, these applications started appearing.
Of course we need faster links into people's homes. Of course under-1Gbit links into homes in 2022 are ridiculous.
10gig is insane.
Finally, can your SSD even keep up with speeds like this?
**Oh and I thought I’d note, it is the positive use of “insane” like “mad good”.
COTS nvme SSD can easily do 1GB/s which easily maxes out non-optimized 10Gbit due to overhead alone. Burst that is for write, but you get the idea.
Meanwhile here in Silicon Valley, the fastest connection I can get at home (sonic.net) is 22Mbps/1.7Mbps. In the office I have Comcast Business which is a bit faster (don't recall exactly) but certainly well below 100Mbps.
So yes 1Gbps sounds quite insane. 10Gbps is out of this universe.
Now, 400G uplinks are still pretty reserved, but 100G (and its' breakout, 4 x 25G) is common as dirt now.
This is quite efficient due to the PCIe4 support not hogging many lanes (half of the 8 lanes can be enough depending on usage), but to use the speed one needs a 25G switch. Is the cheapest way just an Ryzen with 5 of these directly attached to the CPU, and running routing on the CPU via DPDK?
[0]: https://geizhals.eu/qnap-lan-adapter-qxg-25g2sf-cx6-a2609459...
Or what if you want a bunch of historical data for some sort of analysis, and you'd like to start that now, not next week when it finishes downloading? If it's historical (i.e. sitting there available to download) then faster is always better right? If your connection is the bottleneck then all else equal why wouldn't you want it?
(I imagine 10Gbps is about the point currently where it's not necessarily going to help going faster on your end, unless you you have multiple (concurrent) sources, or know that even at that you're still still the bottleneck.)
https://www.esri.com/arcgis-blog/products/product/imagery/ma...
Being from a soon north korea 2.0 to be (russia,) I can say it was my surprise to see Internet being so bad across the developed countries when I was travelling in the previous decade. Canada - ridiculously expensive traffic, Germany - no comments, UK - sometimes good, sometimes 128kbps DSL, USA - 20mbps DocSis everywhere
India, Pakistan, Bangladesh, China, much of Africa had FTTH as the dominant way of home Internet connection for a while.
Just fyi as I know you aren’t a native speaker, it’s ‘chip in’, if you were native I’d assume a typo, probably is for you too, but it’s a phrase easy to mishear and when I was learning a second language I appreciated these corrections.
We moved back to NL and have 1GBit fiber, and there has been a short outage once in three years. I know that there are a still a lot of addresses without fiber, but when I last checked the stats, about 50% of the addresses has the possibility to get a fiber subscription. Heck, even my parents who live in a small rural town have fiber.
I wonder how the upgrade might look considering that 10Gbps hardware is quite expensive (and house cabling might need upgrading) and 2.5Gbps/5Gbps is quite new and hard to find router or laptop dock/hub supporting it.
At my previous flat 150mbps was the fastest available. So it varies a lot in London on where exactly you are.
We are still creating newbuilds in cities without fiber. In other news, someone fucked up construction and left an entire complex of brand new apartment blocks with 7 mbps internet
https://www.ispreview.co.uk/index.php/2017/03/mistake-leaves...
I visited Nepal several years back. They were going straight to wireless. All of the villages in the mountains had a solar panel on their roof that they used to charge their cellphones and there was a cell tower on the ridges.
It was surreal to see villagers use Facebook but have no road access to the main city.
It is easier to upgrade everything in some Blatic states than in some US cities.
In actual fact, our telcos were heavily subsidized during their formative years, granting them a monopoly, rights of way, and helping to pay for their infrastructure. In return for a guaranteed profit margin, we had extreme control over their pricing structure and guarantees of service quality and coverage.
Then we allowed them to be privatized and deregulated, in exchange for which we get fucked. Which is, as far as I’ve ever been able to tell, the inevitable outcome of converting public services to private.
Privatisation is good if competition is fostered at the same time. For example by opening existing physical infrastructure to multiple service providers.
Where I live we had (historically) one fixed lined govt provider. When mobile arrived we got 3 private providers. With fibre we got multiple hardware installers (ie multiple companies digging fibre) and then dozens of ISPs running on that.
So far, so good.
That is the official terminology used in school books required by the official curriculum.
I'm from what used to be a third-world country in the strict sense of term, I'd now refer to my country as a country in the global south.
Meaning what, in the southern hemisphere and not Australia or new Zealand?
We have 195 countries in the world and only 33 of those are entirely on the southern hemisphere.
Here is one example of an illustrative map:
https://i.redd.it/w2tv9dda1xzy.png
Or not necessarily that the inventor of the phrase thinks like that themselves, but rather that they think most people think that way.
The closest geographical designation was the Brandt line drawn in the 80s showing a north and south divide in terms of global wealth distribution [0].
[0] https://en.wikipedia.org/wiki/Global_North_and_Global_South#...
I understand if this comes across as pedantry but the global north and south terminology does a better job at leveling the playing field when talking about wealth and progress disparities.
As per your other question, Australia is part of the global north, China is part of the global south in the global north and south groupings.
I mean... but how? It's just the exact same concepts with less intuitively obvious words. Arguably it frames the disparity as a more immutable part of a nation; it doesn't fit normal language patterns to talk about a country moving north/south.
How? South Korea, Japan, and Singapore are richer than France or Spain now. As nations develop the geographic terminology makes less and less sense. Developing nation is much more accurate.
In other words, effectively developed = what we are, developing = trying to be what we are.
It also places unnecessary upper bound on progress (if you call yourself developed, I guess you can stop developing then?).
Ultimately it is forcing a finite game onto the world.
Global south sounds like a better term.
That’s the plan, yes.
> Ultimately it is forcing a finite game onto the world. Global south sounds like a better term.
Except that it’s geographically inaccurate as well as kind of racist. Japan, Singapore, and Korea can get richer than France or Spain, but they still get lumped in with the other non-white people in “the global south.”
Just to clarify, the plan is to stop developing?
> Japan, Singapore, and Korea can get richer than France or Spain, but they still get lumped in with the other non-white people in “the global south.”
Global south/north divide lumps SK, SG and JP into the north. I have a bigger problem with it putting Russia there, but not e.g. Argentina.
Besides, note how the term "developed" sounds like it has an objective, final and fixed definition, but always invites moving goalposts. Now you use some financial metric. Tomorrow someone uses education levels or democracy.
**
The truth is that everybody should keep developing, and development can entail different things to different people. If you say "I'm developed, and that other person is developing", "we are good, but they are not good yet", all you get is resentment when in fact you want global cooperation.
"Developedness" is a vague, subjective measure that indicates duck all except hubris on the part of whoever coined the term (not coincidentally, that was developed countries).
A vague measure warrants a vague term. "Global South" is far from perfect, but it is less bad. It is obviously geographically inaccurate, which on a meta level is quite fitting.
I don't think I could ever use either term without air quotes, but at least with global south they are more or less implied.
Only if you arbitrarily make exceptions for what’s “south”—ignoring the fact that Singapore and Taiwan and Korea are south of China. The phrase itself implies a permanent association between “south” (I.e. non-white, since skin color is a function of latitude) and “under-undeveloped.” That seems far worse to me than acknowledging that developing countries are indeed trying to become like developed countries.
I'm not against replacing "third world", but something with the word "south" is ambiguous and counterintuitive.
'Southern hemisphere' is fine though I think? And I assume that's what was meant.
Your conclusion is my point.
Doesn't make sense. The second world countries wouldn't drop the third world when realigning to first world.
First world: NATO aligned. Second world: Warsawa pact aligned. Third world: Non-aligned.
Sweden and Swtizerland are traditional third world countries.
>>The term "Third World" arose during the Cold War to define countries that remained non-aligned with either NATO or the Warsaw Pact.
https://en.wikipedia.org/wiki/Third_World
Add in a requirement/desire for it being real-time (or equivalently (you'd get further and further behind) continual) and it could be necessary if large (per second) enough.
After having had to pay Vodafone 50 euros a month for 10Gb of data through LTE for years, due to the impossibility of getting any connection faster than 3Mbps (even as a business), we are also, through the miracles of governmental financial incentives and magnificent planning (only 2 years late), getting the joys of high speed internet in our area. 100 Mbps. Hoooray!
Telecommunication in Germany is a cancerous, price-gouging cartel that needs to be eradicated at all costs, packaged on the first-available Falcon Heavy flight, and propelled at escape velocity out of our solar system.
I'm impressed the Ookla speedtest server can deliver 25Gbit.
The 25gbit network card he mentions costs $400-$500, that's cheaper than I would have guessed.
However, I only have a gigabit router, so most of that isn’t utilized, assuming that’s true for all my neighbors as well (Sonic’s only offering here is 10gbit)
Usually got around 500-600Mbps over WiFi on a fast device.
It's not surprising since is testing against his ISP's test server.
It has gotten better over the years, but even the best available consumer connection is 1000 Mbits down and 200Mbits up.
Majority of the tech Germans I meet are embarassed about the internet and telecom infra, being an advanced economy.
Hopefully the new government fixes things.
I do agree about the general state of German internet connectivity though. However, seems to be catching up quickly, especially in rural areas.
I have not actually upgraded my router at all so I not profiting as much from it as I could.
Still, love init7 and their service.
I don't knwo anything about SSD so maybe 25Gbit/s is easily achievable, just talking from my experience of copying from a USB Drive to my local disk.
Average SSD write speeds are usually around 1GB/sec these days, but most folks could build a RAID of several SSDs and be able to keep up with such a connection, as well.
I would also assume that most folks who would bother to install 25Gbps, or anything close, probably have a lot of devices in their homes to take advantage of it at once.
Personally, I find it very hard to saturate the 1Gbps link I have. Most servers won't push more than around 200Mbps out to you, and I'm basically never doing 5 tasks at once that are that bandwidth-intensive. But it's nice to have some overhead, just for fun.
https://www.servethehome.com/exploring-best-zfs-zil-slog-ssd... is several years old but still a good read.
To fully utilize the connection you can use storage backed by memory. For example ESXi allows you to back a vm by a "virtual pmem" disk that uses a chunk of your host system dram
My 2 fiber connections are 500mb/s and each cost $350 to activate, $150/mo each not counting the costs for static IP's. Trenching it in was $3k. I'm more than happy with 500mb but the price could be lower. For my specific use cases it is less about bandwidth and more about latency/peering arrangements. I would personally be happy splitting as low as 100mb/s across many devices with fq_codel or cake and cdg on my little firewalls. I can't really complain though. The alternatives here are 4G LTE or Starlink.
Absolute madness. But this kind of bandwidth isn’t meant for a single machine.
Having fiber without pon is great, means no new infrastructure for decades. But anything above 10gbits will be too much to use realistically, even for smaller companies.
Individually no one person/port can use more than a gigabit and can't saturate the switch in this case, but combined they could utilize far more.
WiFi goes up to 1Gb/s, if you're lucky. Sure, some WiFi-6 APs have a 2.5Gb/s connector, but that's not what you want or need, unless you're a high-density enterprise. WiFi-6E will possibly improve that a bit, but it will take WiFi-8 to get anywhere close to saturation.
Wired, you can do 10Gb/s for server systems, which are, amongst other things very loud and not very suitable for placement anywhere near humans. 2.5Gb/s support is spotty, and 1Gb/s still the only thing that works reliably.
So, exactly which residential application requires 25Gb/s is not very clear. Yes, it's cool, but not very useful, and faulting manufacturers (especially in times of crippling supply-chain limitations) for not fully supporting it is questionable.
2.5Gb/s can easily be done with a lot of laptops and workstations these days; 10 Gb/s isn't quite there yet (and 25, 40 and 100Gb/s are definitely in the server-only fiber-or-DAC-only realm)
It's also easy to open and replace the fan with something less noisy. Just remove a few screws, it's a standard size with a normal connector on it.
There's also big external heatsink so you could rig up a big, slow, fan to help it out a bit.
If you want to be green by the way, use fiber/DAC. The 10G copper SFP+ modules are power hungry, and Mikrotik recommends not placing them next to each other. Also the extra power draw is likely to result in fan use if you have a lot of them.
The Mikrotik ones (S+RJ10) are based on Marvel chip and they are the more power hungry / run over 30m only variety. On the other hand, they can negotiate 2.5G or 5G if necessary and support a proprietary protocol to tell the switch about it, so you will that in SFP+ properties.
As I have written in the sibling comment, I have good experience with BCM84891-based transceivers. CRS305 can handle two (still not next to each other, obviously).
Mikrotik CRS305 (4xSFP+) and CRS309 (8xSFP+) are both passively cooled. They are pretty much silent :) though the blue led takes some tape to be less shiny.
> 10 Gb/s isn't quite there yet
If you really, really need RJ45, look into BCM84891-based transceivers. They still get hot, but not as much as others (according to datasheet, takes 1.6W at 30m and 2W at 80m). I also managed to get stable 10g over 20m Cat5e with them.
There are a few quiet 10 GbE switches, the passive Mikrotik ones others are mentioning, but this one is a quiet actively cooled one if you want RJ45: https://www.qnap.com/en-us/product/qsw-1208-8c (Lots of combo RJ45/SFP+ ports)
Consumers around the globe have had increasingly common access to 1 Gbit/s for a decade and there still aren't any other great, common use cases for it beyond very high quality video streaming.
It didn't take very long for computer use to need more than 640K by comparison. In the computer realm those edges were being constantly pushed at that time. Such is not the case with bleeding edge consumer broadband speeds today.
10 Gbps is still below the 7 GB/s that a single NVMe on PCIe 4 (which is readily available) can achieve.
25 Gbps is still below that.
I'd say 100 Gbps is where the current practical maximum is more or less. You'd have a hard time writing or uploading data that fast on anything resembling consumer hardware.
"Everybody believes quotes on Internet" - Abrahamo Lincolni
The bigger issue with 25G is that it’s well into the range where any OS’s default TCP settings won’t provide anywhere near line speed, and once you solve that it can expose other non-network bottlenecks. I have dual-SFP28 cards in both my workstation and NAS, both connected to the network via a10G Mikrotik switch and directly to each other via a 25G point-to-point link. Now after all that tuning I have a fast network but run into the SAS bus bottleneck for any file transfers exceeding the size of my app take write cache :p
SR is ~$40 and LR is ~$60, so more expensive than SFP+ modules but not out of the reach of someone who even has a capable SFP28 port.
The last time I thought I have not enough BW on my 50Mbps connection I learned what it's just my T440 is not enough to decode h265 extremely compressed stream over an SFTP streaming over WiFi, which gave around 20 Mbps at best. After toying|fighting around with wireless settings I found the main culprit was my Intel 7260 (or whatever) WNIC, not the CPU, Internet connection BW or the server throughput (10Gbit, despite being an IIS instance).
YMMV
My 10gbaset switch is all of 23 dB when the fan kicks on and I have other switches with sfp+ that don't even have a fan.
Granted, it's beyond the definition of overkill (but I wouldn't mind it!)