Amazon's Project Kuiper satellites nail online orders from orbit

Amazon is boasting a 100 percent success rate for its Protoflight mission, having demonstrated that a pair of prototype Project Kuiper satellites are capable of streaming video, conducting two-way video calls, and buying stuff on Amazon. Following a recent successful test of the satellites' thrusters, Amazon confirmed that …

[elsergiovolador]

Excitement

Regarding customers, Amazon claims its standard terminal can deliver speeds up to 400 Mbps and can be produced for less than $400. A compact, portable terminal will reach 100 Mbps, while a larger enterprise device will hit 1 Gbps.

That would be exciting 20 years ago.

What is the point in deploying these legacy speeds today?

Also do they have a plan how to take these satellites down or littering the space is fine?

[Joe Gurman]

Re: Excitement

Because there are many, many people in the world who have no access to such bandwidth.

[MachDiamond]

Re: Excitement

"Because there are many, many people in the world who have no access to such bandwidth."

Don't forget to mention that those same people can't afford to pay for that bandwidth with many not able to afford a computer to access the internet at all. There are existing satellite internet providers that cover the regions of the world where average earnings are high enough to afford it from Geosynchronous orbit. The advantage (just about the only one) to LEO sats is ping time. Tribes living a traditional lifestyle in the Amazon rain forest can't afford and don't need internet to begin with and certainly won't be playing Call of Duty if they decide to break with the old ways and start embracing tech.

[Flocke Kroes]

Re: Excitement

In part you are right about much of the less connected parts of the world not having the money for one terminal per family. The most obvious target market is rural US which combines high income with customer hostile lousy service. Airlines and ships will pay real money for world wide connectivity. From there each terminal must be shared between a larger and larger number of people to become cost effective. Internet connections for the poor have shown some value. Some farmers were in a better negotiating position after they were able to access international pricing for their crops.

LEO has other giant advantages over GEO. The shorter distance allows more bandwidth at less power. The small antenna size creates a wide beam that drastically limits the number of GEO satellites that can share frequencies. It is like switching from the Vatican's old 250kW transmitter sending one signal to the whole of Europe to 10,000 smaller local radio stations.

The economics work out for mass produced satellites and cheap launch. Kuiper does not have either at this time but Bezos has shown the required commitment to haemorrhage money for years to take control of a large market.

I am glad that there will be a choice of LEO ISPs even if it is a choice between a crocodile and a shark.

[MachDiamond]

Re: Excitement

"LEO has other giant advantages over GEO. The shorter distance allows more bandwidth at less power. "

Ok, but Viassat's constellation is three birds and Starlink's build out has been quoted by Elon and Gwynne at 42,000 with a 5 year life span. At the five year mark, it will take launching 70 replacement sats every other day forever to keep the full constellation in service.

[Anonymous Coward]

Re: Excitement

> "Don't forget to mention that those same people can't afford to pay for that bandwidth with many not able to afford a computer to access the internet at all."

Cliché. But wrong.

Coverage gap - that's the technical term - is as follows (% of pops):

- East Asia and Pacific 6.9%

- Europe and Central Asia: 7.3%

- Latin America and Caribbean: 5.8%

- Middle East and North Africa: 7.3%

- North America: 4.1%

- South Asia: 8.1%

- Sub Saharan Africa: 12.1%

It's just about how expensive it is to draw backhaul fibres. At some point Non Terrestrial Networks (NTNs) are a better proposition (hence all the standards from Release 14 onwards).

And I'm not even citing the Usage Gap figures.

[Justthefacts]

Re: Excitement

Brazil actually has 165 million internet users, who also spend longer glued to the internet than the EU.

[MachDiamond]

Re: Excitement

"Brazil actually has 165 million internet users, who also spend longer glued to the internet than the EU."

I'll accept that. Now, how many of those are in major cities where there is wired infrastructure? Outside of the largest cities, what's the average annual family income?

A village can club together and get one terminal at a central location, but the question becomes if that adds any value to the people in that area.

[Version 1.0]

Re: Excitement

The society is unique - Brazilians normally work to get everything workable for everyone they are working with, most other countries normally only work to make money from the users. It's just a different society attitude in Brazil ... a country that I have been sent to work in for about 30 years now and it's great!

[elsergiovolador]

Re: Excitement

How is that an excuse for delivering legacy product?

[John Brown (no body)]

Re: Excitement

That would be many of the same reasons that WiFi speeds lag wired, such as penetration through obstacles. From orbit, the atmosphere is an obstacle, not to mention clouds and other weather phenomena such as snow and rain.

If you want 10Gb/s or more from orbit, you can have it. But it will cost exponentially more to get the much larger sats up there and each will have to serve fewer users thus each user bearing more of the launch and running costs.

Todays nearly successful Starship launch may address these issues in a year or three, but don't hold your breath. The consumer ground stations will have to be more powerful and therefore more expensive to transmit back to the sats too.

[Anonymous Coward]

The UPLINK is the problem

Remember your phone's emission power is limited around 2W (to avoid frying your brain).

With these 2W, it's already a challenge to reach a BTS (Base Transceiver Station) 10km away.

This is why the uplink in Apple-Globalstar case is L-Band (1-2Ghz) as opposed to S-Band (2-4Ghz) for downlink.

Also, there will always be less satellites in the sky than BTS's on the ground. Yet BTS's have to schedule uplink f-t resources in their own tiny area of coverage to support many uplink sessions.

For uplink to work, you need low frequency and therefore you can support less throughput. But you cover much more than the terrestrial cell, so that throughput is shared amongst many more users.

In short, the physics says that LEO networks can only be a complement to cellular networks. They are decently well adapted to downlink intensive use cases, but will never be able to perform as well as cellular terrestrial networks in uplink demanding use cases. Existing sat phones have different antennas and are a tiny fraction of the installed base.

HAPS's could be better technical options.

[MachDiamond]

Re: Excitement

" Todays nearly successful Starship launch may address these issues in a year or three, but don't hold your breath."

To date, SpaceX has only had a single Starship test article NOT explode (SN15).

The flight looked much better as all of the engines lit and stayed on during boost, but I'm very curious to see if it was the booster's FTS that blew it up or if the little bang from the engines ruptured the body in between the tanks. If it was FTS, that was a major change from what they implemented last time. The Starship FTS activating seems more likely after a first look. Something happened to the LOx supply and it could have flamed out too early and FTS destroyed the airframe since the mission was over at that point and missing an impact with Africa was important. I'm only throwing out a really rough guess. The fuel consumption in Starship seemed high. Eventually, the craft is supposed to re-land propulsively and the indicators were getting really low with what is only a shell and not a completely kitted out interior. Lots of heat tiles were shedding too. Re-entry would have been fun.

Maybe by tomorrow somebody will post some overflight video of the launch pad. Elon was saying in an interview in May that with all rocket engines going, the expectation is to erode 12cm/s of steel from the OML horizontal surfaces. 60cm of steel eroded in 5 Seconds! Yikes.

[Flocke Kroes]

Re: is littering space fine?

These days you must have a plan to prevent littering to get a launch license. Failure to implement the plan results in a fine.

[pidloops]

Re: Excitement

Because it's about 5x faster than I can get here in rural Arizona at any price. I would LOVE 100 MBps.

[xyz]

Mmmmmm....

Hate to say it but I can stream video, whilst GF rabbits away on Teams, whilst my mum is wittering stuff to me via whatsapp and if I had peace and quiet I could order something from Amazon at the same time. 65€ a month via Starlink. (>300mbps up, around 40mbps down, 20ms or so latency)....

And no ads for Prime or constant special offers**

** pretty much guaranteed as part of the Amazon in space T&Cs.

[John Brown (no body)]

Re: Mmmmmm....

"** pretty much guaranteed as part of the Amazon in space T&Cs."

Good point! With Amazon as your ISP, tey will have total control over your Internet "experience", not least the ability to inject their own ads. Anyone remember the early days of free 0800 dial-up in the UK? At least some provided their own hacked browser which enforced banner ads along the bottom or top of the screen inside the browser window but outside of the "web" windows. I never looked but I assume they were delivering webpages into some sort of frame/iframe to keep their own ads outside of the users control so unblockable. And those "free" PCs that were available, possibly only in the US, which required a net connection and enforced the watching of adverts every so often.

Neither lasted long or were especially successful, but with that previous knowledge and Amazons current abilities and knowledge (and users acceptance of shit like this being at a much higher tolerance level nowadays), I could easily see Amazon doing advertising/hijacking on their own end to end service.

[Crypto Monad]

Re: Mmmmmm....

You forget that there is competition. Starlink is well-established, and Amazon will have enough trouble catching up on the customer base without actively modify the content.

Amazon *could* offer a reduced monthly rental in return for accepting injected adverts, as indeed could any other ISP (terrestrial or otherwise), but it's been tried and failed.

More likely they'll offer Prime Video streaming which doesn't count towards your monthly bandwidth cap, or something like that.

[Pete Sdev]

Re: Mmmmmm....

I imagine that, at least for EU customers, there'll be a with-ads tariff and a more expensive without-ads tariff.

Similar to what Amazon already does with it's Fire tablets.

[MachDiamond]

Re: Mmmmmm....

"Hate to say it but I can stream video, whilst GF rabbits away on Teams, whilst my mum is wittering stuff to me via whatsapp and if I had peace and quiet I could order something from Amazon at the same time. 65€ a month via Starlink. (>300mbps up, around 40mbps down, 20ms or so latency)...."

I've got the telco coming next week to install 500MBS up and down fibre for $50/month. The unreliable cable company is getting the heave-ho for 115down/12up that turns on and off like a yo-yo in the summer when the pole mounted line amps overheat. I'm in the desert and they've painted the stupid things black. The latest bill from the cableco is $85 and seems to rise every 6 months. There's also a private wireless internet company in town if you know who to talk to and the guy that used to own the Radio Shack installs and sells Hughesnet and Viaasat. For being out in the hinterlands, we are getting spoiled for choice. Verizon hasn't yet brought out home internet over 5G for our town, but it's marked as "coming soon". I expect that T-mobile and AT&T will follow on behind. How much was the cost to get into Starlink?

[Kermit de Frog]

Shiny tin foil, Miss P.

[TheMaskedMan]

"Anyone remember the early days of free 0800 dial-up in the UK? At least some provided their own hacked browser which enforced banner ads along the bottom or top of the screen inside the browser window but outside of the "web" windows."

I do remember, though I never used one or saw one in use. Rather than an iframe, I assumed that they would use something like the adware companies of the day - a component dropped on a form (for Delphi, anyway) that was basically a separate ad-fetching browser.

That was easy to implement from the developer's point of view and was actually a really good idea in theory. It meant you could give software away for free and still make money from it. Totally ruined by the ad companies getting so greedy that it - quite rightly - became a menace to be removed on sight.