It’s not really meant to compete with cable/fiber. This is for people stuck with geo stationary trash, dsl, microwave relay, etc. It’s a massive market and can take SpaceX to a trillion dollar company easily without ever stealing a cable customer.

The incredible thing is that despite not having to compete with wired connections, it is almost competitive with them.

They likely have to price it a bit above wired connections just to keep load manageable in dense areas, but it puts an upper bound on how badly monopolist incumbents can treat their customers.

While I don't think it will actually replace cable/fiber, I think it will drive prices down and service quality/customer treatment up, so its existence is a huge benefit even for people in populated areas that will never actually use it.

What is the latency? I find there is a material difference in my internet experience with 5ms fiber connection vs 100ms+ cable connection. I imagine it’s worse for satellite links due to the limitations of physics.

20 ms was cited elsewhere in the thread. Unlike older satellite Internet solutions that use geostationary satellites (~36000 km altitude), this uses satellites in low earth orbit (~550 km altitude, with additional, slightly lower and higher levels planned).

That would be a welcome surprise if it delivers and finally gives people an alternative to the one wired option they usually have.

Their 12000 satellites will have a total bandwidth of 240000 Gbps. I don't see how they get to a trillion dollars on that. A 100 Mbps stream being uploaded to one satellite and then relayed through others and finally download to and end user is going to use at least 200 Mbps of that total system bandwidth.

That puts an upper limit of 1.2 million worldwide simultaneous 100 Mbps streams, and that's probably being too generous. In a given region the limit is much smaller. The numbers I saw said 500k simultaneous 100 Mbps streams in the US.

Overselling can push those numbers up, but I don't think they can oversell enough to take them to a trillion dollar company.

> Overselling can push those numbers up, but I don't think they can oversell enough to take them to a trillion dollar company.

You’re vastly underestimating how much you oversell as a consumer ISP. Fiber companies with dedicated lines to your house can’t even handle all of their customers pulling 100mbps at a time because the bottlenecks are the IXPs.

The AMS IXP is one of the largest exchanges in the world, and it only hits about 10,000 gbps max per day.

You’re also forgetting that $100/mo is a cost offered to consumers on the ground. Commercial shipping and military won’t bat an eye at paying $1000/mo for a measly 10mbps to a ship in the pacific. Oil rigs, mines, military bases, etc, etc.

Great insight, and I agree. With a lot of family and some friends in rural parts of the U.S., the weak internet and cable connections are detrimental to any sort of internet-based tech.

>trillion dollar company easily

trillion dollar??

it seems WAYYYYYYYY too much for

...or with shitty Cox

People who don't live adjacent to a public thoroughfare are sometimes quoted 5 or 6 digit installation costs for cable.

https://community.spiceworks.com/topic/1607004-charter-cost-...

https://www.reddit.com/r/Comcast/comments/9ytg25/comcast_wan...

https://www.dslreports.com/forum/r9512648-TWC-TWC-wants-want...

https://arstechnica.com/information-technology/2015/08/want-...

This is for those people.

Hi! I live on a major arterial in a major US city. We have buried utilities to my house, and my house was never wired for anything faster than what you can get on a phone line. Even though all six of my property-line adjacent neighbors (whose houses are in stone-throwing distance) have gigabit connections.

After 9 months and a call to the City Council, I finally got a quote from Comcast: $37,000 for copper. They laughed when I asked what it would cost for fiber.

Just run ethernet to a neighbor?

It's a competitive price for a satellite service. HughesNet and Viasat will charge upwards of $100/mo for ~25 Mbps service with significantly higher latency and limited data allowances (<100 GB/mo).

For people who live in rural areas this is pretty typical.

Typical in rural areas is $100 for 5mbps or even less. This is less than a mile outside of a town with a university and decent cable connection — they just don’t lay cable out into the country, and the phone lines can’t (or won’t) handle higher speeds

I would expect it'll be relatively the same cost everywhere, but I could be wrong in that. These satellite (and point to point) internet services aren't really designed for people who have a lot of internet options.

$150 for 50Mbps is what I'm paying for Comcast Business, and they haven't been able to render an invoice for months.

I pay 70/month (introductory pricing) for 100mpbs, not that big of a difference and that is with a data cap.

What's the currrent word on data caps? Unlimited for now?

Unlimited for now but they've hinted they could cap them later.

If I was them, I'd offer uncapped Internet, but prioritize based on past bandwidth usage if there is congestion.

So if you use a lot of bandwidth, you still get full speed at night, but during peak times, other users get priority and your connection becomes slower.

Bonus points if they only take bandwidth usage during peaks into account, incentivizing customers to move bulk transfers to off-peak hours when the traffic is effectively free.