Hacker News Reader

Comment by GMoromisato | Hacker News Reader

I don't think the math works.

There are 23 million rural homes in the US and about 3 million miles of rural public roads. Let's say you wired just the public rural roads (ignore going from the road to the house).

It costs $30,000 per mile to put up aerial wiring. $60,000 per mile underground. So we're already at $90 billion for wired poles and $180 billion for underground. And that's just for the wires--we're not including any of the switches and routers for actual internet.

By comparison, the Starlink system cost about an order of magnitude less ($10 billion).

View on HN

I appreciate you actually taking a moment to think through the cost, but I think we could start with some pragmatism and look to run wires to people who are within a reasonable range of existing systems, of which there are many.

Clearly not every public road needs wiring. Then, consider that you could run wired connections to wireless access points to increase high speed wireless coverage. 1 wire to light up dozens of homes in areas which currently have no service beyond DSL.

Doesn't Starlink have some annual upkeep costs? Maybe $1-2 billion per year to replace aging satellites?

Yes, but Starlink needs to exist for military, planes, boats and other essential very rural services as well. So the upkeep should pay for itself.

And of course Starlink has to be for the whole planet, so just comparing it to the US would be a false analysis.

Of course you also need to upkeep the physical infrastructure. Specially if you don't put all those lines underground.

But one would need to do some more real work to compare. I would also say that a real program for urban fiber makes a lot of sense in more places. But I would love to see somebody take a shot at this, what would be the best if you started from 0 today?

You don't think aerial lines have upkeep costs? ice, tree branches, hurricanes, etc

Would 5G towers be a better alternative?

Reply on news.ycombinator.com