Comment by dawnerd
6 months ago
Hey are absolutely not replacing “legacy” isps and certainly not mobile. Even if they had perfect coverage, sat signals are way too sensitive to obstructions.
6 months ago
Hey are absolutely not replacing “legacy” isps and certainly not mobile. Even if they had perfect coverage, sat signals are way too sensitive to obstructions.
No, but they are replacing bad ISPs. I have a relative in Brussels, while there is 10gig fiber on a nearby street, he's stuck on 100/10 coax, and to add insult to injury, Starlink is cheaper.
Coax is an old tech, but it is surprisingly innovative and pushed limits a lot with right equipment. Newest full duplex and extended spectrum models could potentially reach 10/10 Gbps and all they require is changing some passive splitters in the cable plant and RPD plus CM supporting new modulation. Which are way way cheaper than satellites.
What I'm saying, is as soon as there is a real competitor pressure, ISPs can upgrade their deployments in under a year or two, even without touching buried copper. Of course they can also choose not to do that too :) .
I mean your relative is maybe a member of the tech elite who needs amazing bandwith but 100 Mbps/10Mbps is not going to be limiting for most people. Coax is already pretty fast considering it probably takes its source from fiber at street level and mostly constrained in uploading. I just went from coax to fiber and I cannot tell the difference when browsing, streaming or sharing. Maybe it is because my devices are stuck on wifi 5 but even then I have my doubts.
On the other hand : "Starlink users typically experience download speeds between 45 and 280 Mbps, with a majority of users experiencing speeds over 100 Mbps. Upload speeds are typically between 10 and 30 Mbps."
That doesn't sound meaningfully different. What is the price difference ?
You are quite right. Also in practice Starlink has random jitter and packet loss at unpredictable times, very visible when talking to my colleagues in Ukraine when they are on backups or in the country. It's fine solution, but landlines are for now superior. Also Starlink's bandwidth depends a lot on the majority of people staying on the landlines. Starlink is nothing short of miracle, but it has limitations. Interesting to see the if the v2 and v3 will upend the status quo.
cable is still more stable than starlink. I have regularly issues on a teams call with starlink while it just works with cable.
And come on 100/10 is not bad despite the other 10gig fiber
True. They're replacing legacy ISPs in areas where your choices are high latency geostationary satellite service, dialup, or DSL where the nearest DSLAM is far enough away that it may as well be dialup.
And density is limited. Starlink cannot more than a tiny fraction of a city.
Limited by the amount of satellite coverage? Or on the ground limitations from ever greater division of bandwidth? I guess those two are directly correlated, but they could in theory add microwave relays into the mix to help offset higher population dense areas from lower density ones with less used bandwidth.
I think it's mainly a limitation of the beam steering: terminals that are close together need to be seperated in bandwidth so that they don't intefere with each other, and there's only so much bandwidth available. Tighter beam steering could help but that is something that tends to bump up against physical limitations.
(and it's an area where it would need orders of magnitude improvements to address the density of cities, it's not really close at the moment)
3 replies →
If anything they'll go for the lucrative customers that _need_ a signal to go faster through vacuum than through glass.
Maybe some decent revenue offering sat to cell for the traditional carriers.
There is no workload that latency sensitive is there? If it is you just move the server where it's needed. I.e. you do HFT next to the stock exchange.
Definitely not a usecase for Starlink.
For the microseconds-chasers, there's microwave relay links, say between Chicago and New York (ref e.g. https://bullseye.ac/blog/economics/inside-the-world-of-high-...). Sending a signal up a few hundred km and down again a few hundred km adds way too much latency, and signal-hopping between fast-moving satellites adds way too much jitter for "such applications".
[dead]
They do make internet convenient and somewhat reliable for places that had no other options (on my boat, etc.) But I doubt they can displace wired or microwave relays, and why would they? On-earth systems are much cheaper to maintain even if the cost to LEO reduces by another order of magnitude.
To be fair, in Tokyo I see a lot of ISPs pushing 5g routers. Many buildings have fiberoptics pulled to the basement and then use VDSL for the last meters, and I bet they'd rather move everyone over to 5G than have to start actually installing proper fiberoptic internet. In Norway, 5g has been advertised as something groundbreaking and radical. We have been told "now surgery is finally possible with mobile networks" (hospitals don't have fiberoptics??) and similar. Very Apple 2010s "The ipad can now be used by (good person) to do (good thing)"-like. But nobody cares, real users don't see any benefit.
A normal person will probably never notice the difference between 4g and 5g because of what they use their phone for, and giving every household a proper fiberoptic line is probably a much better quality of life improvement. But ISPs dont want that future. They want everyone to be connected to these neighborhood hubs that don't require last-100m-cables and expensive construction. The same can probably be said for Starlink. It's "Good enough", and that's good enough to get sales. They don't care about the quality of the product they deliver, or if fiberoptics are superior. They care about sales.