Live data from Hacker News

SpaceX plans worldwide satellite Internet with low latency, gigabit speed

arstechnica.com

151–160 of 540 posts

Re: SpaceX plans worldwide satellite Internet with low latency, gigabit speed

#151
Try reading to the third paragraph before asking your "honest question":

> SpaceX wants to launch 4,425 satellites into low-Earth orbits, with altitudes ranging from 715 miles to 823 miles. By contrast, the existing HughesNet satellite network has an altitude of 22,000 miles.

That's about — basic arithmetic alert! — three percent the distance for the SpaceX satellites compared to existing geosynchronous ones.

Re: SpaceX plans worldwide satellite Internet with low latency, gigabit speed

#152
Something I find interesting about this: what does it mean for countries who monitor and filter/block content (think China)? Presumably since this isn't using their national infrastructure they would lose all control of it, no? I wonder how those countries would respond to SpaceX wanting to let their citizens use the service. Unless I'm missing something it seems to me like they wouldn't really be able to stop it.

Re: SpaceX plans worldwide satellite Internet with low latency, gigabit speed

#153
post #22

I don't understand the math. 800 satellites at 23 gbps seems to be able to provide 1 gbps to 18,400 connections at any given time. This seems to be many magnitudes off from what would be needed to have any impact. What am I missing?

People aren't generally using a 1 gbps connection at 100% capacity 100% of the time. If we assume customers use an average of 1 TB of data per month and the time they use it is pretty spread out (seems reasonable since they are going to be spread out accross timezones eventually), you can serve about 6 million customers with 800 satelites @ 23 GBPS, and with 4400 @ 23 GBPS you could do ~33 million. Admittedly these n…

Using the 4400 example. People pay ~20$ per month for under 10GB cellular data. So 100GB for 150 million globally people globally seems viable. If that 150 million is paying 15$ per month that's 30 billion per year. Optionally, I suspect a few customers would pay a lot more for a larger monthly cap.

The issue is the antenna needs to be around 2 foot by 2 foot.

But, with autonomous cars people may want significant bandwidth in there cars. An hour or two of video a day seems reasonable especially if they can cache popular content.

Honestly, if they can get the price down this could be very profitable.

Re: SpaceX plans worldwide satellite Internet with low latency, gigabit speed

#154
post #64

Earlier quoted context omitted.

This still requires ground stations. Though I guess you could pay for "censor resistant" traffic where your data is bounced among the satellites until someone has a US ground station in sight...

That still wouldn’t give you access to piracy sites, for which the FBI has required US ISPs to censor them from their DNS servers. I’ve had this occur quite a few times now that 8.8.8.8 (yes, verified with DNSSEC to be the real one) and DNS servers of US ISPs returned different IPs than the authoritative nameservers, and the returned IPs always showing a neat "The FBI has seized this property due to piracy" link.

> he FBI has required US ISPs to censor them from their DNS servers.

That's a pretty serious statement, the government threatens infrastructure providers all the time but I haven't heard of the FBI requiring ISP level censorship of DNS.

> I’ve had this occur quite a few times now that 8.8.8.8 (yes, verified with DNSSEC to be the real one) and DNS servers of US ISPs returned different IPs than the authoritative nameservers

Can you give some examples?

Re: SpaceX plans worldwide satellite Internet with low latency, gigabit speed

#155
post #45

Earlier quoted context omitted.

Doubt it. Elon's made his fortune by replacing existing products with superior alternatives. He's said in every case that driving cost down is one of the most important parts of creating something people want. Why do you think he wouldn't do the same here? He knows everyone hates existing telecoms. He's trying to create a better alternative.

> He knows everyone hates existing telecoms. Not true everywhere. I'm paying < $40/month for a 100 MBit connection and I don't hate my telecom.

I'll hate on Comcast enough for two people, then. Or ten, or a thousand. It's unbounded, really.

Re: SpaceX plans worldwide satellite Internet with low latency, gigabit speed

#156
post #64

Earlier quoted context omitted.

This still requires ground stations. Though I guess you could pay for "censor resistant" traffic where your data is bounced among the satellites until someone has a US ground station in sight...

That still wouldn’t give you access to piracy sites, for which the FBI has required US ISPs to censor them from their DNS servers. I’ve had this occur quite a few times now that 8.8.8.8 (yes, verified with DNSSEC to be the real one) and DNS servers of US ISPs returned different IPs than the authoritative nameservers, and the returned IPs always showing a neat "The FBI has seized this property due to piracy" link.

You were probably witnessing normal DNS propagation after the authoritative name servers had been changed. Keep in mind that the nameserver data you see in "whois" may not be up to date. When a site is seized, law enforcement goes direct to the registry (GLTD operator, such as Verisign), not the registrar.

See http://domainincite.com/2766-icann-had-no-role-in-seizing-to... for an example

Re: SpaceX plans worldwide satellite Internet with low latency, gigabit speed

#157

Such a genius move, surprising nobody saw it coming. Just think: he will already have access to a large base of customer's roofs (which SolarCity effectively owns) --- 1 million by 2018. Then, it's just a hop, skip, and a jump to an on-demand, fully-cached, mesh network. They're probably not far from being able to provide internet service to all SolarCity customers, just a bit of a modification to the contract.

Free internet access for Tesla vehicles too, I imagine.

Re: SpaceX plans worldwide satellite Internet with low latency, gigabit speed

#158

Earlier quoted context omitted.

Where are those numbers from? I get a consistent 12ms ping with my cable internet (200Mbps) to local cloud servers, and have seen as low as 1-4ms on fiber.

Yeah, I'm pretty sure they're made up, and they lack context. 40ms for ADSL... From where to where are you pinging? Gamers will regularly achieve ~20ms pings to other gamers on the same continent. Where as pinging around the other side of world (under-sea cables) is going to add approximately 150-200 ms. Also FTTH vs. ADSL vs. Cable really shouldn't make much difference with regards to ping (they're all fibre for the…

You should try Australia. Undersea cables definitely adds more than 150-200ms (but I guess you mean the Atlantic) to everything here.

Re: SpaceX plans worldwide satellite Internet with low latency, gigabit speed

#159

Earlier quoted context omitted.

Competition makes a big difference. Just over a year ago with Google Fiber rearing up, my ISP offered 150Mbps for $35. This year with Google Fiber downsizing, the big American ISPs are back to their duopoly: 25Mbps for $40 or 10Mbps for $30. Hopefully SpaceX is able to force them to lower rates for everyone.

It's not going to replace internet for cities/towns. Think of the latency! It is directed towards middle of nowhere areas where nobody provides fiber.

If the satellites are in LEO, latency won't be that bad. At 700km height mentioned elsewhere in the thread, added latency might be 10-20ms based on halving the numbers at https://www.dslreports.com/forum/r100277-Whats-the-latency-p... (from 16 years ago).

Re: SpaceX plans worldwide satellite Internet with low latency, gigabit speed

#160
napkin math (chemistry style): (4425 satellites * 23Gbps) / ((60,000PB of fixed internet traffic / 1 month [1]) * (12 months / 365 days) * (1 day / 24 hours ) * (1 hour / 60 minutes) * (1 minute / 60 seconds) * (8 bits / 1 byte) * (1024 Tera / 1 Peta) * (1024 Tera / 1 Giga)) =

An astounding O(54% of global internet traffic) [2]! Obviously there are many many more variables in this [3]. Nonetheless this will be super interesting to follow.

[1]: my estimated current data usage from https://en.wikipedia.org/wiki/Internet_traffic#Global_Intern...

[2]: http://www.wolframalpha.com/input/?i=(4425+*+23)+%2F+((60,00... (1024+%2F+1)+*+(1024+%2F+1))

[3]: Side note, someone should create a multi user algorithm website (it might be able to be done over google spreadsheets) but adding in extra factors like revenue streams and what not sounds like it would be fun to play around with.

Post reply on HN