I'm also skeptical. A single LTE node can have a total bandwidth of 1+ Gbps. In Poland there are probably over 20k such nodes. Yet, the average speed you get is about 25 Mbps.
With Starlink we're talking about covering the whole world, not a 40 million country. So I believe it may be usable for rich people in rural areas, but in densely populated areas it stands no chance to LTE (and 5G is coming and promises lower latency than physically possible with Starlink).
> I'm also skeptical. A single LTE node can have a total bandwidth of 1+ Gbps. In Poland there are probably over 20k such nodes. Yet, the average speed you get is about 25 Mbps.
I imagine usage is not evenly distributed between those nodes!
Also LTE connections are limited by the speed of the connection running to the base station. even worse, if that base station is on a microwave link to another tower that then has fiber running to it, expect slower speeds.
Also there is the phone hardware. Take Wi-Fi for a minute, which is a more controlled environment, the performance difference between different chipsets rated at the same speed can be huge! Firmware, drivers, and just how the chip is wired up. Plenty of opportunities to screw things up, sometimes even with the same chipset between different laptop manufacturers.
And finally, there is going to be a power trade off, I wonder how inefficient those LTE chips are when running full bore! Now I am curious as to what that graph looks like. :) Some sort ∩ shape I imagine where it starts being more efficient as you download faster so the radio is on for a shorter period of time, and then getting less efficient as you try to squeeze every last bit of performance out of the chipset.
With Starlink we're talking about covering the whole world, not a 40 million country. So I believe it may be usable for rich people in rural areas, but in densely populated areas it stands no chance to LTE (and 5G is coming and promises lower latency than physically possible with Starlink).