Hacker News Reader

Comment by kilroy123 | Hacker News Reader

While I mostly agree with you, as I too have traveled a lot (I've been to all 7 continents). I've been to many places with very poor internet.

Still, I think you're missing one very important point. The likely biggest user of this system will be the US military and Department of Defense. Just how they use the iridium constellation, only phone calls aren't enough in the 2020 and beyond.

They need the same system but with broadband speed _internet_. They can not be dependent on some other nations infrastructure and there are a lot of boots out at a sea and places where there isn't a cell tower.

The math just doesn't add up. Mark my words, the StarLink system will never work or be profitable without the US Gov being a major customer.

View on HN

About the DoD, I think Starlink will be beneficial as UAVs will be able to stream with less latency and higher quality, which eventually means better targeted democracy.

"targeted democracy", my new favourite phrase of the day

> The math just doesn't add up.

Can you demonstrate that math in a comprehensive manner? I've seen this kind of claim repeated frequently and it's never supported.

> The likely biggest user of this system will be the US military and Department of Defense.

The biggest user of Starlink by an extreme margin will be consumers. It won't be a close comparison.

There will be at least a million consumers eventually using Starlink around the world. The network has capacity to handle that globally. The military is going to push that much data through Starlink? No they're not. A million consumers pushing HD video streams through it will rather comically outweigh anything the military will be doing.

From a business perspective, what matters is $'s, not Gbps.

I'm pretty sure the US military will be paying more $'s than any other country's people.

And it'll be used for being able to fly drones with weapons to kill anyone in the world in 30 mins. With that ability, you can decommission a lot of aircraft carriers and spend that cash on decent connectivity for your drone fleets...

I'm also curious about specific math.

But I think the DoD billing would be different than you describe. Specifically, even if they're not pushing GBs of video all the time, they'd be paying a lot for guaranteed bandwidth at any time. Pure speculation on my side though - mostly based on seeing overinflated requirement docs.

There is some math on the Internet that's commonly linked. With some super pessimistic assumptions, it shows Starlink just barely breaking even (aka, not losing money, but not making much either). Most of those assumptions have already been proven to be false. Doing the math with less pessimistic assumptions shows it to be insanely profitable. I'd do it here, but this has already dropped off the front page so nobody will see it.

One customer I weirdly haven't seen mentioned is high frequency traders. Shouldn't Starlink's latency be lower than that of cables?

would that help them though? i mean it appears to me they are already as close as possible to avoid latency and use microwave links and stuff so that is pretty hard to beat i guess. the only thing i can easily imagine being beneficial is getting information across the globe faster but then i lack the skills to see how one profits from that, ie. being closer would still be better, no?

Arbitration. If you can see an asset being traded cheaper in Australia than in the US 10ms before everybody else does - you basically pocket the difference at 0 risk.

Mark Handley has done a bunch of videos analysing this on his channel[0].

The TLDR is that there is a lot of room for Starlink to beat existing terrestrial options IF they can get the laser links working on-orbit. This video[1] shows estimates using what was known about proposed configurations in 2018

[0] https://www.youtube.com/channel/UC-k1j7M2-hBfXeECd9YAQ_g

[1] https://youtu.be/QEIUdMiColU?t=237

How would they be faster than microwave / laser terrestial towers, that already travel around the great arc?

You can't really cross oceans with microwaves though. So a London from can gain a competitive advantage against other London firms on Wall Street exchange.

It’s interesting. At this kind of distance, I wonder how topologies of cables and satellite constellations compare.

Cables run relatively straight under the sea. If they are optic fibre, they run at a lower speed than light in vacuum.

Satellites are in space, so communicate pretty much at speed of light, but probably not in a straight line, depends where the satellites are.

HFTs are indeed great at gaming such things, will be interesting to see what happens.

Probably not too useful for high speed traders as they can afford and are technically sophisticated enough to relocate their systems closer to the exchange, but I know that enterprise software that suffers heavily due to latency in APAC regions. Lower-latency connectivity from those regions to EU/US datacenters will be a small game-changer for many businesses.

Reply on news.ycombinator.com