4 ms·

Really surprised to see the beating heart of tech that is HN turn into a luddite forum on this occasion. Last week I got upgraded from 4G to 5G for home broadb

Really surprised to see the beating heart of tech that is HN turn into a luddite forum on this occasion. Last week I got upgraded from 4G to 5G for home broadband here in London, UK. It just became commercially available in my area (on Three) As I said in another comment, my average speeds went up from 5-10Mbps to 150-250Mbps and peak ones from 20-25Mbps to 350-400Mbps. It absolutely is the game changer it's advertised to be in terms of enabling new real-world use cases. Remember how video streaming was theoretically possible on 3G but totally shit in practice until 4G came along ? 5G is a similar kind of step up. You can expect an order magnitude more bandwidth on mobile than with 4G, and that matters. In other words : the hype is real.

povertyworld 7y ago
Is data going to be cheaper too? Otherwise, it's not going to change anything for most people. It'll just mean I can blow through my monthly quota in a day instead of a week.
- renaudg 7y ago
  This particular service doesn't have a quota, it's unlimited.
manmal 7y ago
My 4G connection in Austria has been able to do 120Mbps for years now, sometimes even 180Mbps. What you are seeing might simply be that you are mostly alone in your cell, and not really 5G related. Once more people have a 5G modem, things will get worse for you again.
- 420codebro 7y ago
  That was true with 3G to 4G, yes. But with 4G to 5G there is hope for relief in this area. One of the big features of 5G is the ability to handle a magnitude order more density with more "intelligent" frequency allocation, modulation, and (the big one) - LoS beamsteering. Of course, with all things in life there are tradeoffs. The tradeoff in this case is a 5G provider infrastructure radio every few hundred feet in urban areas to enable the LoS beamforming with the density factor required.
- creato 7y ago
  By design, 5G uses many more and smaller antennas, so it solves exactly the problem you are pointing at.
davb 7y ago
I had a similar experience (also on Three) when 4G first rolled out. Blazing speed, low latency, low contention. Unfortunately as the technology became more mainstream, the speeds dropped and peak contention increased. A small scale roll-out is absolutely not indicative of real-world performance. In Glasgow, I'm getting around 5Mbps on Three over 4G. In Japan and South Korea I regularly got far in excess of 100Mbps over 4G. It's more about backhaul bandwidth and capacity planning than radio technology. 4G is more capable than you'd think, and any speed real-world increase we see with 5G is almost certainly down to better network design. Edit: autocorrect typo - "tech logo" to "technology"
- renaudg 7y ago
  In the particular case of Three there's a simpler explanation and it means things will be very different this time : they completely lost the 4G spectrum auction war to competitors (EE and Vodafone mainly), but they didn't repeat that mistake with 5G : they're the only carrier with >100Mhz of spectrum and are likely to dominate performance comparisons for a while.
  - johnnycab 7y ago
    Your comments contain fundamental errors. I won't go into great detail, why Three failed to acquire 4G spectrum and how it was subsequently awarded significantly valuable frequency from EE in the process. However, there is a lot more to how Three (Broadband) came to acquire frequency in the 3.4─3.8GHz spectrum. There was no auction neither were there any wars. It was via the acquisition of UK Broadband & Relish, which was already a part of PCCW. It was originally awarded the spectrum, contingent upon certain stipulations ie. building, maintaining and continued investment in telecoms infrastructure. The current 5G product might bear the moniker, but it is still using the infrastructure put in place by UKBG Relish in parts of Central & Greater London. The speeds are also dependent upon relevant upgrades to the masts, operator equipment, backhaul etc., more importantly the UE has to be up-to-date to take advantage of any speed boosts. It will still not count as real 5G for sometime yet, as there are standards which are still being ratified and other players have yet to join the game. It would be naive to assume that Vodafone, the biggest operator by spectrum holdings, followed by BT/EE, will just defer their incumbent position, without bidding next year! https://www.ispreview.co.uk/index.php/2019/08/three-uk-aim-for-fastest-mobile-broadband-crown-as-5g-goes-live.html https://www.ispreview.co.uk/index.php/2019/08/three-uk-aim-f... https://www.capacitymedia.com/articles/3659839/Three-UK-agrees-to-buy-PCCW-owned-UK-Broadband-Limited-in-250m-deal https://www.capacitymedia.com/articles/3659839/Three-UK-agre...
com2kid 7y ago
Just ran a speed test, 55Mbps down, 5.5Mbps up. An increase in upload speed would be a game changer. That download speed can stream 4k w/o issue. The upload speed could use some improvements, I can imagine better scenarios if I could stream high res video off to a cloud AI processing system!
superkuh 7y ago
>Remember how video streaming was theoretically possible on 3G but totally shit in practice until 4G came along ? 5G is a similar kind of step up. You can expect an order magnitude more bandwidth on mobile than with 4G, and that matters. 5G-NR standard only gives a 15-20% gain in data thoroughput over the same bandwidth as compared to 4G LTE. The vast majority of the increase in thoroughput has to come from the millimeter wave bands. Think about streetlights to get an idea of coverage and propagation. Now think about running internet fiber to every streetlight. It's quite a task. It'll only be done (and has only been done so far in the USA) in areas that require high bits/khz/km^2 like entertainment stadiums and parts of the largest cities. Additionally, almost no two countries actually have overlapping 5G bands. It's a mess of competing standards. So it'll be different frontend modules for controlling the antenna arrays required on the handsets (high power usage and space requirements) due to the integration required. Handsets are going to have to get bigger. Unlikably big and specific to region. And beyond that the 5G setups being implemented now aren't stand alone. They're both LTE and Non-Standalone (NSA) 5G. The 5G just provides extra bandwidth and isn't used as a control for the handsets. It more than doubles the complexity of setups because of the trouble RF designers have to take with harmonics and intermodulation. All the competing national freq band standards mean different filtering solutions. The future will not be evenly distributed.

povertyworld 7y ago

Is data going to be cheaper too? Otherwise, it's not going to change anything for most people. It'll just mean I can blow through my monthly quota in a day instead of a week.

renaudg 7y ago
This particular service doesn't have a quota, it's unlimited.

manmal 7y ago

My 4G connection in Austria has been able to do 120Mbps for years now, sometimes even 180Mbps. What you are seeing might simply be that you are mostly alone in your cell, and not really 5G related. Once more people have a 5G modem, things will get worse for you again.

420codebro 7y ago
That was true with 3G to 4G, yes. But with 4G to 5G there is hope for relief in this area. One of the big features of 5G is the ability to handle a magnitude order more density with more "intelligent" frequency allocation, modulation, and (the big one) - LoS beamsteering. Of course, with all things in life there are tradeoffs. The tradeoff in this case is a 5G provider infrastructure radio every few hundred feet in urban areas to enable the LoS beamforming with the density factor required.
creato 7y ago
By design, 5G uses many more and smaller antennas, so it solves exactly the problem you are pointing at.

davb 7y ago

I had a similar experience (also on Three) when 4G first rolled out. Blazing speed, low latency, low contention. Unfortunately as the technology became more mainstream, the speeds dropped and peak contention increased. A small scale roll-out is absolutely not indicative of real-world performance. In Glasgow, I'm getting around 5Mbps on Three over 4G. In Japan and South Korea I regularly got far in excess of 100Mbps over 4G. It's more about backhaul bandwidth and capacity planning than radio technology. 4G is more capable than you'd think, and any speed real-world increase we see with 5G is almost certainly down to better network design. Edit: autocorrect typo - "tech logo" to "technology"

renaudg 7y ago
In the particular case of Three there's a simpler explanation and it means things will be very different this time : they completely lost the 4G spectrum auction war to competitors (EE and Vodafone mainly), but they didn't repeat that mistake with 5G : they're the only carrier with >100Mhz of spectrum and are likely to dominate performance comparisons for a while.
- johnnycab 7y ago
  Your comments contain fundamental errors. I won't go into great detail, why Three failed to acquire 4G spectrum and how it was subsequently awarded significantly valuable frequency from EE in the process. However, there is a lot more to how Three (Broadband) came to acquire frequency in the 3.4─3.8GHz spectrum. There was no auction neither were there any wars. It was via the acquisition of UK Broadband & Relish, which was already a part of PCCW. It was originally awarded the spectrum, contingent upon certain stipulations ie. building, maintaining and continued investment in telecoms infrastructure. The current 5G product might bear the moniker, but it is still using the infrastructure put in place by UKBG Relish in parts of Central & Greater London. The speeds are also dependent upon relevant upgrades to the masts, operator equipment, backhaul etc., more importantly the UE has to be up-to-date to take advantage of any speed boosts. It will still not count as real 5G for sometime yet, as there are standards which are still being ratified and other players have yet to join the game. It would be naive to assume that Vodafone, the biggest operator by spectrum holdings, followed by BT/EE, will just defer their incumbent position, without bidding next year! https://www.ispreview.co.uk/index.php/2019/08/three-uk-aim-for-fastest-mobile-broadband-crown-as-5g-goes-live.html https://www.ispreview.co.uk/index.php/2019/08/three-uk-aim-f... https://www.capacitymedia.com/articles/3659839/Three-UK-agrees-to-buy-PCCW-owned-UK-Broadband-Limited-in-250m-deal https://www.capacitymedia.com/articles/3659839/Three-UK-agre...

com2kid 7y ago

Just ran a speed test, 55Mbps down, 5.5Mbps up. An increase in upload speed would be a game changer. That download speed can stream 4k w/o issue. The upload speed could use some improvements, I can imagine better scenarios if I could stream high res video off to a cloud AI processing system!

superkuh 7y ago

>Remember how video streaming was theoretically possible on 3G but totally shit in practice until 4G came along ? 5G is a similar kind of step up. You can expect an order magnitude more bandwidth on mobile than with 4G, and that matters. 5G-NR standard only gives a 15-20% gain in data thoroughput over the same bandwidth as compared to 4G LTE. The vast majority of the increase in thoroughput has to come from the millimeter wave bands. Think about streetlights to get an idea of coverage and propagation. Now think about running internet fiber to every streetlight. It's quite a task. It'll only be done (and has only been done so far in the USA) in areas that require high bits/khz/km^2 like entertainment stadiums and parts of the largest cities. Additionally, almost no two countries actually have overlapping 5G bands. It's a mess of competing standards. So it'll be different frontend modules for controlling the antenna arrays required on the handsets (high power usage and space requirements) due to the integration required. Handsets are going to have to get bigger. Unlikably big and specific to region. And beyond that the 5G setups being implemented now aren't stand alone. They're both LTE and Non-Standalone (NSA) 5G. The 5G just provides extra bandwidth and isn't used as a control for the handsets. It more than doubles the complexity of setups because of the trouble RF designers have to take with harmonics and intermodulation. All the competing national freq band standards mean different filtering solutions. The future will not be evenly distributed.