Starlink broadband speeds slow as subscriber numbers grow Elon Musk's Starlink satellite broadband service has seen a decline in download speeds around the world as more and more subscribers sign up, perhaps making the company a victim of its own success. The data comes in a report by network Intelligence company Ookla, which states it has monitored the performance of satellite …
This is exactly what many people on El Reg and elsewhere predicted.......
Just like any other service were there is contention, it is great in the initial start-up then as number increase, available bandwidth decreases. StarLink can keep lobbing satellites up but eventually they will run out of space (literally).
Over at the Starlink subreddit there are a couple of typical complaints:
1) Starlink needs to ship gear faster, I've been on the waitlist for months!
2) Starlink needs to ship gear slower, my speeds keep decreasing!
I went to Starlink from 18 Mbps down 0.8 Mbps up DSL. Starlink, even with increasing congestion, is an upgrade for me. If you have FTTH, or even a good cable provider, it's not going to be an upgrade for you.
I'm hoping Starlink allows legacy carriers to abandon remote areas, and forces them to upgrade areas like mine. Small cities, exurban areas, and similar medium density areas have fallen through the cracks (metro areas tend to have lots of options, certain rural areas get subsidized FTTH, I fall inbetween both)
I think all those ISPs are waiting until it has become "common knowledge" that paying 99$ for average broadband even in urban regions is now "normal". Then they'll start shipping 5G fixed wireless at a lower price than Starlink, increasing their current median prices significantly.
I am convinced that 5G mmWave Fixed Wireless is being kept back deliberately due to this thinking. All the frequencies have been allocated long ago.
T-Mobile already does. 5-G fixed mobile broadband for $50/mo. No contract. I have both Comcast & FIOS into my house (load balanced) for business reasons. Both are cheaper or equal in price to the T-Mobile deal, but there's a 5G tower near me. It will be interesting to see which one raises prices first causing me to substitute T-Mobile for them.
As for Starlink, I used to be only able to get 1.5Mbps DSL into my Vermont cabin, until after 18 months on Starlink's waiting list, I got Startlink. It usually measures out around 60Mbps down, 12 Mbps up, which is a vast improvement. But it does drop out occasionally despite my having a clear view of the sky. Since the DLS is only $24/mo, and my router can handle two ISP's, the DSL is always on to handle those Starlink glitch times.
How everyone can dance around the obvious part of this.
The constellation was planned to have many, many more satellites. They haven't been flying new ones at the pace they projected, as they are waiting on the lower cost of their heavier lift launch capability. That's why this is happening. So yes, as you add customers in a given area, eventually the speeds slow down. And at this point no one is reporting that that slowdown is crippling, or below service levels.
Because the team that runs this aren't total incompetents and are aware of their metrics. Because the customer signup process literally is designed to handle this. Guess what? If you live in a major uban center and tried to sing up and it told you there was a wait list, that's why.
So yeah, love it or hate it, Starlink growth is going to be limited in areas that are already saturated with enough subscribers to strain the network, at least until the MKII constellation starts up. They will also be bigger, allow less dense constellations, and have a reduced albedo to make photoshopping them out of the pics from your telescope easier. None of this is new information, how is not being mentioned?
That's great if the servers you want to connect to are sited on an oil tanker but doesn't help much with connecting to the other 99.99% of the internet. It's not good enough to just get the packets down to sea level. They need to actually reach shore and the rest of the internet ;)
Or 5-10m off HugesNet for almost as much money(and a bigger dish probably).
Elon's ego aside, the utility of this greatly exceeds it downsides. The incumbents had little incentive to upgrade their service as a duopoly. They let the tech in their constellations stagnate for years, and had no plans to scale up capacity to handle a large number of users.
Yeah, there will be some impacts, and it's not replacing land based cellular data anytime soon, but it's adverse effects are manageable, and the benefits will have a positive global impact. Whole nations currently rely on a single link, at much lower speeds, from a single provider. One blip, especially in a disaster, and the whole population is cut off from the world, and potentially needed aid.
I wouldn't rush to cut those service lines just jet(the amazing boy wonders network isn't mature enough yet), but I expect Starlink terminals are going to continue be popular in isolated locations all over the world.
Starlink is already shaking things up in the USA (the only country where there is a big enough market for Starlink to be worthwhile building). The terrestrial incumbents are already moving to protect their interests.
One such example is, despite 5G spectrum having already been allocated but not fully deployed, they're asking for more spectrum in a band that just so happens to overlap Starlink's frequency of operation. There is currently a court case about this, the terrestrial companies saying that it won't interfere (see some dodgy report they have had written that uses an inappropriate antenna model to say it won't interfere), but it's a deeply suspicious move.
What this is doing is beginning to show just how uncompetitive the US market is, even to the extent of anticompetitive behaviour involving legalised electronic warfare...
It's a mess. For the US Gov, they need Starlink to succeed because they need, for geopolitical reasons, Starship and superheavy booster to work.
222mbps down and 30mbps up and price has been reduced to 70 euros a month. I'm in Spain.
My GF streamed in HD, that dead queen thing for over 7 hours on Monday and there was one 1s drop out over the whole period.
Given that a Starlink groundstation seems to jump on any ISP's bandwidth I'd be looking at throttling going on at that end.
While that sounds impressive, it's important to remember that streaming (especially on demand) is very tolerant of latency / jitter and dropouts. It may be buffering 10+ seconds in advance, which means that your 1 second dropout was actually closer to 11 consecutive seconds of actual connection problems. Any problems shorter than 5 seconds or so would be hidden.
Why are people in urban areas with access to wired internet signing up for Starlink? Is it because they're Musk devotees?
Is the internet in the US of A so utterly crap that even in major urban centres (eg. Los Angeles) people would rather pay for Starlink than plump for a wired connection? I understand the appeal out in the sticks, in the middle of nowhere where there's no other alternative, but when there is an alternative WTF? No wonder there's contention....
The nodes in the constellation are designed to fall back to Earth at the end of their useful life, which is claimed (by SpaceX, yes SpaceX) to be 5-7 years... this means the entire constellation will have to be renewed every 5-7 years... This is mental - it's going to cost over $10Bn to loft the entire constellation (up to 42,000 nodes at approx $250K/node) and then having to rinse & repeat every 5-7 years. And then there's the fixed costs for the ground stations, uplinks, dishes, staff etc.
I just can't see Starlink ever being profitable - at a minimum, it is looking at costs of $2Bn/year just to maintain the space network.
It just makes me think that Starlink is nothing but a very clever way for Musk to syphon money into SpaceX which itself will go under without the business from Starlink (and Government grants/tax breaks).
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
