[myrmidon]

For reference: Starlink is <$20bn yearly revenue (a bit over $1k/customer).

Unless it manages to massively eat into mobile network market share (pretty unlikely in my view) growing past $100bn (yearly revenue) or so seems unrealistic to me.

Even with cheaper launch costs, it is not clear to me that Starlink would ever be interested in offering service for like $80/year (=> price competitive with mobile carriers in low income nations).

[nomel]

> Unless it manages to massively eat into mobile network market share

Several mobile carriers are already using them for backend of remote towers.

[AntiUSAbah]

For something relevant? Why would they do this? Dark fibers is an easy thing and you need to put power to these remote towers anyway.

[nomel]

You need a cell tower every 20 miles or so, less for a performant one. There's not dark fiber at that density in remote regions (remote is, by definition, lacking infrastructure). There's a better chance of power, and for the cases there's not, solar/battery is fine for a cell tower, especially one designed for lower power remote installations.

[AntiUSAbah]

The initial cost for a dark fibre on a power pool is very low and can give you low latency and high bandwidth.

Starlink antenna needs a lot more energy and has a high montly cost.

That shouldn't be very economically for long.

[AntiUSAbah]

And to add to this: They still have packet loss when they switch the satelites every 1-4 minutes.

[nomel]

Packet loss is an assumption with cellular, so accommodations are built into all stages, including the variable rate vocoders used! And, a bit of packet loss is still better than literally no service, which was the previously, economically driven, state for these places.

There's a 20% increase in satellites planned, with possible 300% increase in the future, so it should reduce.

[fragmede]

If only there was some sort of transmission control protocol that was resilient to packet loss!