SpaceX Starlink Targets Near Global Service by 2021

SpaceX Starlink is initially providing service to the northern United States and southern Canada. SpaceX states that by 2021 the company will provide service to the populated world.

Customers receive internet from the satellites in space via a user dish terminal and Wi-Fi router device. SpaceX sent and email to some potential beta customer describing the 19-inch dish terminal with mounting tripod and router cost $499 Upfront cost and the installation process.

Connection speeds are 50-150 mbps now and have 20-40 milliseconds of latency. By mid-2021, latency should be down to 16-19 milliseconds.

SOURCES- Michael Sheetz, Starlink, SpaceX
Written By Brian Wang,

22 thoughts on “SpaceX Starlink Targets Near Global Service by 2021”

  1. You are talking about the number of active customers, which is usually 1/20 of the number of subscribers according to the traditional telco metrics.

  2. Since the service is very local due to low satellites and smart antennas beams, they can't say they are servicing one country while actually providing costumers at other counties.

    Well, full sized countries like USA, Canada, Russia etc. I agree, it's not going to work.
    Tiny, "fun-sized" countryettes like in most of Europe, the Middle East etc. sure you can.

  3. Can compete with Comcast in my area. I would have switched in a heartbeat before WOW came on as a serious (and better) competitor. If WOW pisses me off enough (e.g. by raising the price), I might still switch.

  4. I think you really can think about this as cells defined as areas which can communicate with only one satellite at a time. Obviously they are constantly orbiting and handing off every 90 seconds or so (if there are 60 satellites in an orbital shell with 90 minute orbital period). The number of customers in this cell cannot exceed the capacity of a satellite. It's not really a mesh network at this level of granularity as by definition there is only one satellite that can be communicated with and no redundancy.

  5. No, they have to operate on a basis of signal strength and number of connections allowable since the satellites are not individually serving a fixed position, they are in LEO and constantly moving at high speed orbiting every 90min or so. They have to be able to pass the connection off to the next satellite in range like a mesh network.

  6. I'm not sure this is a good way to estimate the potential capacity. I think you need to determine the cell size that is served by one satellite at a time, and the most customers you can have in that area is the capacity of one satellite. Adding more satellites allows you to shrink the cells (more orbital shells for narrower cells widthwise and and more satellites per orbital shell for shorter cells lengthwise). The cells are naturally smaller closer to equator and larger further away.

  7. The big thing is that they will likely need base stations in each country (at least until satellite to satellite communication works, and perhaps even beyond for capacity). Not sure how much a country could do to disrupt a foreign ISP from beaming in service other than to be a nuisance from a payments perspective.

  8. Caveat, by 2021 it will only handle 1.5-3 million subscribers at best since that's about 2000-2500 satellites that can handle 5000-10000 simultaneous connections. If they hit the 4k sat mark and get 10k connections per satellite, they get maybe 4-5 million subscribers.

  9. Don't get your hopes up. At least the big dictator country has their citizens pinned down pretty effectively. In that country it would be super easy, barely an inconvenience, to check that no one has a Starlink antenna…

    And if someone would be caught, it would be prison for him or have his social points reduced to the point where he can neither travel nor can his kids be enrolled into college…

  10. Agreed – but if Starlink licenses local companies to 'redistribute' service using local wireless, the end-user cost of internet could be a lot less, and the end user equipment could be as cheap as free for any computing device with WiFi built in – a phone, tablet or laptop.

  11. I'd love to see a black market in antennas in countries that interfere with free communication. Imagine how easy it would be to hide one of these, and how hard to trace the narrow beam back to the source, particularly if it were moved regularly. One under a fiberglass camper on a pickup would be nearly impossible to find.

  12. I'd guess the service will eventually be cheaper than incumbent providers in most of the US. There may need to be a limit to subscription in densely populated areas.
    Likely, it's the end of 2021 that is targeted for rollout, but prices won't compete with cable internet until starship allows quick population of the constellation, and higher orbit laser "backbone" sats are available.
    Once it's rolling, starlink will print money.

  13. $50 USD a month with $500 upfront is a lot compared with metropolitan ISPs.

    It's very cheap compared with existing satellite Internet, and with better bandwidth/latency.

    But again, it' s a lot for people in developing countries.

  14. From a business perspective, I think the biggest question for SpaceX is how they intend to market this. I.e. how are they going to get their customers? The language issues alone will be a huge impediment to global marketing – and frankly I don't see SpaceX wanting to go that direction.

    In developed nations, they could go through existing wireless ISPs, offering a way for them to offer improved service at the same cost.

    In less developed nations, they might need to create or help create national companies that do something similar – e.g. setting up local area WiFi services that take care of all the marketing and target local customer needs.

  15. Yeah, the ownership structure will be interesting. Many countries have foreign ownership limits on telecom companies, so Starlink mothership may need to license technology and access to global satellite network o local subsidiaries that are either floated and majority owned by locals or joint ventures.

  16. Is Starlink part of SpaceX or is it its own company?

    In any case, Musk certainly will have to eventually make some choices about what he wants to focus on and lead more closely.

  17. Mr. Musk needs to cash out of Tesla and ride the ISP gravy train all the way to Mars. No reason Starlink can't make SpaceX $10 billion per year.

Comments are closed.