>>13045679A 4K60 video streaming in realtime peaks at around 110Mbps or 13.5MiB/s. Streaming songs on spotify and doing an F5 reload of this page did not exceed 0.5MiB/s.
I've got a gigabit connection, because its cheap for me. But the point is that, I don't ever use 99% of the bandwidth continuously, only in infrequent bursts (watching content, watching anime, etc). Even gaming doesn't use more than 10-25Mbps of bandwidth.
Having 28Tbps means they can support arguably 5x that number; around 1.4M people; and this lines up because they for their FCC filing for ground terminals and dishes, wrote in their licensing doc that they intended to make around 1M of these as of ~6 months ago. Since then, they've launched a lot more satellites and the capacity has gone up. So assuming an extra 400k is reasonable.
Finally, like others have said; oversubscribing is very common for ISPs. Theoretically you could double that number to 2.8M total subscribers as potential; but let's assume that grants SpaceX an extra 100k subscribers to 1.5M total @ 1400 satellites.
That means:
- 1,500,000 x $80/mo = $120M/mo | I'm assuming pricing will drop by ~$20 to 79.99 or 80 bucks for live as opposed to the beta which $99 + $499 for the dish (which is currently sold at a loss)
- $120M x 12 months = $1.4Bn in revenue
- Assuming 12 F9 flights a year for Starlink @ $40M in launch costs = $480M in overhead, literally
- $1.4Bn - $0.48Bn = $920M/yr
That's pretty good for only 1.5M per people subscribed. If you add a zero at the end of the number of subscribers, Starlink revenue goes from $920M to $9.2Bn; add another zero: 1.5M -> 15M -> 150M (world wide), Starlink's revenue goes from $920M -> $9.2Bn -> $92Bn
I'm jealous of every SpaceX employee that has shares for Starlink. Being able to print ~$100Bn/yr at full network deployment, is so insane, for the life of each satellite across 12k satellites, amortized over 5 years, means that Starlink prints $500Bn/5yr. They can literally BUY Amazon.
