Starlink IPO

[Cosmacelf]

So, bear with me here, but I'm thinking of revenue potential beyond just home ISP. I'm sure this has been mentioned before in a Starlink thread (and is probably why Elon gets peppered by the banking types on putting starlink in cars). In theory, it's possible to shrink a satellite transceiver enough to fit into something just a little bit larger than a cell phone. It would be a device thicker than a cell phone, but still smaller than current Sat phones. This would open up 2nd and 3rd world countries with poor quality or expensive cellular systems to a voice and internet service from Starlink. Eventually we could be looking at IoT devices (the cell phone receptor in my SunPower system SUCKS - just as a quick thought) that are outside of covered cellular areas.

Starlink is something that should have a good number of traditional industries (ISPs, rural cell providers, alarm companies that charge outrageous rates for wireless connectivity, etc.) scared shitless.

Do you have any backup for your contention of shrinking antenna sizes that small? If not, then what's the point of speculating. I mean, I can imagine lots of things, doesn't mean they are possible.

 

[mongo]

So, bear with me here, but I'm thinking of revenue potential beyond just home ISP. I'm sure this has been mentioned before in a Starlink thread (and is probably why Elon gets peppered by the banking types on putting starlink in cars). In theory, it's possible to shrink a satellite transceiver enough to fit into something just a little bit larger than a cell phone. It would be a device thicker than a cell phone, but still smaller than current Sat phones. This would open up 2nd and 3rd world countries with poor quality or expensive cellular systems to a voice and internet service from Starlink. Eventually we could be looking at IoT devices (the cell phone receptor in my SunPower system SUCKS - just as a quick thought) that are outside of covered cellular areas.

Starlink is something that should have a good number of traditional industries (ISPs, rural cell providers, alarm companies that charge outrageous rates for wireless connectivity, etc.) scared shitless.

Shrinking doesn't work for the base station antenna. You have beam width and power issues going too small, plus you need the steered phase array to track the satellite you are communicating with.
For less developed areas, they can package Starlink with solar and batteries to make a 24/7 hot spot. That can then connect to local wired infrastructure or provide wi-fi/ cellular plus recharging capability to handhelds.

 

[bkp_duke]

Do you have any backup for your contention of shrinking antenna sizes that small? If not, then what's the point of speculating. I mean, I can imagine lots of things, doesn't mean they are possible.

Aside from running operations for 6 different datacenters on two continents, yes. We do utilize satellite as OOB (out of band) connectivity in datacenters for emergency access purposes (hell, I kid you not when I say we still have a few 56k modems for the same purpose).

The transceivers are currently built for satellites in higher orbits with moderate signal strengths. As you cut that distance by several fold using low orbit satellites, even if you reduce the satellite signal strength, you don't need as powerful a transceiver or as large an antenna to catch and maintain a signal reliably. The more satellites in the constellation, the more effective this is because you can potentially multiplex the signal from them and don't have to track a single satellite like you do with current providers (granted, they are in stationary orbit).

 

[bkp_duke]

Shrinking doesn't work for the base station antenna. You have beam width and power issues going too small, plus you need the steered phase array to track the satellite you are communicating with.
For less developed areas, they can package Starlink with solar and batteries to make a 24/7 hot spot. That can then connect to local wired infrastructure or provide wi-fi/ cellular plus recharging capability to handhelds.

Yes and no. Larger antennas, all other things remaining equal, will capture signal more reliably, however your comparison is a bit off because you are comparing capturing signal from a single geostationary satellite (current sat internet) to basically a blanket of them over the sky (Starlink). You don't have to track an individual satellite as long as they are communicating together (this is what we have been told they will do).

I do completely agree, however, that a single large receiver that can capture more signal should be faster and more reliable and would be a fantastic setup to setup a localized hot spot.

 

[mongo]

Yes and no. Larger antennas, all other things remaining equal, will capture signal more reliably, however your comparison is a bit off because you are comparing capturing signal from a single geostationary satellite (current sat internet) to basically a blanket of them over the sky (Starlink). You don't have to track an individual satellite as long as they are communicating together (this is what we have been told they will do).

I do completely agree, however, that a single large receiver that can capture more signal should be faster and more reliable and would be a fantastic setup to setup a localized hot spot.

Power falls off as beam width times distance. Unless you boost the power of both transmitters, you're not going to get good signal strength. Alternatively, if the sats had much larger antennas, that would compensate for low ground station power. A beam wide enough to guarantee coverage of at least one satellite is going to be very weak. Not to mention the additional spectrum use/ overlap and need to select the signal from multiple sats.

 

[bkp_duke]

Power falls off as beam width times distance. Unless you boost the power of both transmitters, you're not going to get good signal strength. Alternatively, if the sats had much larger antennas, that would compensate for low ground station power. A beam wide enough to guarantee coverage of at least one satellite is going to be very weak. Not to mention the additional spectrum use/ overlap and need to select the signal from multiple sats.

We'll just have to see how SpaceX plays this out. A little birdie over there told me a few things and I'm excited to see how well they work in the field.

 

[BioSehnsucht]

Yes and no. Larger antennas, all other things remaining equal, will capture signal more reliably, however your comparison is a bit off because you are comparing capturing signal from a single geostationary satellite (current sat internet) to basically a blanket of them over the sky (Starlink). You don't have to track an individual satellite as long as they are communicating together (this is what we have been told they will do).

I do completely agree, however, that a single large receiver that can capture more signal should be faster and more reliable and would be a fantastic setup to setup a localized hot spot.

Earth to Satellite transmission will use a steered beam, so there is a minimum size for the antenna based on power, area, etc, to meet all these transmission objectives. And for reception from satellite to Earth, the satellite will steer a spot beam (which will actually be quite larger by the time it hits the ground, true in the opposite direction as well) towards the ground terminal and the signal will be received, but reception is not something that can be "steered" in the same way transmission is, so the antenna needs to be large enough to pick up the signal clearly even when it arrives off axis by some 35 degrees or so.

Both ends must track the other for transmission purposes, hence the beam steering with phased arrays, but this is "trivial" (relatively speaking, it's a solved problem), and is helped by the fact the beam is relatively large compared to the originating device by the time it reaches the other end (so if you're off by a little bit in your beam angle, you'll still hit the target). But since this is using phased arrays, there's no physical movement involved same as geosynchronous satellite internet.

Oddly, it was mentioned that the user terminals would have some motors for aiming the antenna, I suppose to idiot proof the installation by having it self-level (or possibly even automatically adjust to be non-level to avoid obstructions to find a clear patch of sky), rather than depending on the installer to be able to use a bubble level or otherwise intelligently aim it. To me this adds unnecessary cost to the user terminal but if they're being priced in the $100's then the additional cost may not really matter much. These motors won't be used to actively steer the beam by moving the antenna, but just to set it up for an optimal view of the sky, whichever direction that may be.

 

[ggr]

cut that distance by several fold

The difference between Clarke orbit and Starlink orbit is a factor of around 50. Inverse square law means the link budget is 2500 times better.

 

[sparcs]

A separate SpaceX official confirmed Shotwell’s statement, saying that SpaceX is considering taking the Starlink project public, potentially in the next several years

SpaceX considers Starlink spinoff as teams ready for mid-February launch – Spaceflight Now

Maybe they are trying to get the first phase of the satellite network up and running to show off to investors.

 

[generalenthu]

I looked a bit into VSAT. their numbers looked pretty ugly, but it's hard to know the business without getting under the surface.

Interestingly, management has been buying, as recently as couple of months ago, in good size. Also Baupost group (Seth Klarman) is a huge stake holder with over 30%. I think this requires a bit more digging before saying this goose is cooked.

 

[Cosmacelf]

I looked a bit into VSAT. their numbers looked pretty ugly, but it's hard to know the business without getting under the surface.

Interestingly, management has been buying, as recently as couple of months ago, in good size. Also Baupost group (Seth Klarman) is a huge stake holder with over 30%. I think this requires a bit more digging before saying this goose is cooked.

Yeah, that's what is confusing me. Bandwidth is bandwidth. In theory, Starlink could replace a lot of the geo telecom satellites, and do it cheaper and better. We've been talking about consumer Internet here, but Starlink could replace traditional geo satellites.

Is it simply a case that traditional satellite telecom has no idea that a freight train is about to run them over?

 

[bkp_duke]

Yeah, that's what is confusing me. Bandwidth is bandwidth. In theory, Starlink could replace a lot of the geo telecom satellites, and do it cheaper and better. We've been talking about consumer Internet here, but Starlink could replace traditional geo satellites.

Is it simply a case that traditional satellite telecom has no idea that a freight train is about to run them over?

I suspect it is just like the traditional automotives. They really REALLY underestimate Elon that much.

 

[UnknownSoldier]

Starlink is really interesting because it really does seem like a massively scaled expansion of the concept originally introduced by Iridium. Anyone remember that? The Iridium 66 satellite constellation is still operating today.

Iridium satellite constellation - Wikipedia

I really think Starlink will be vastly superior to terrestrial cellular telephone networks for phone calls assuming that people will implement VoIP technology over it's network. This is probably a bigger potential market than the Internet access.

 

[HumanPilot]

Revenue and earnings are hard to guess, maybe $300/customer/year and 20 million customers (but it might be 200 million customers).

I'd like to add another number here, because maybe this is not just about consumers but also businesses. I live in Germany and a company I work for just got the proposal for a symmetric internet connection, 100 MBit/s down/up, flat-rate (best effort, no service levels). The one-time installation costs 1,900 EUR (2,080 USD), and monthly fee is 1,400 EUR (1,530 USD) before rebates are applied. All prices without sales tax / VAT. Minimum contract period of 3 years. It isn't a remote location, it's directly in the area that has been upgraded a year ago. And the customer has to wait for around half a year for some civil engineering to connect the customer with the local gateway (paid for by the provider, though).

I don't know their strategy, but maybe they will also cover businesses and there is a lot more money to be made in that field. Above numbers translate to $18k/customer/year.

 

[Matias]

SpaceX considering spinning off and IPO'ing their Starlink business.

Bloomberg - Are you a robot?

Valuation will probably be astronomical

 

[Matias]

Hmm, so that force arms length pricing with SpaceX so basically competitors would need the same pricing. But, do competitors have cheap satellites?

Yep. When SpaceX and Starlink are two different companies with different owners, could SpaceX sell launch services to Starlink with different prices than to Starlink’s competitors?

And also, if owner A owns SpaceX and owner B owns Starlink, why would owner A want to subside owners B businesses even if he could.

 

[mongo]

Yep. When SpaceX and Starlink are two different companies with different owners, could SpaceX sell launch services to Starlink with different prices than to Starlink’s competitors?

And also, if owner A owns SpaceX and owner B owns Starlink, why would owner A want to subside owners B businesses even if he could.

You say owners as if SpaceX could not have a large stake in Starlink.

As to pricing, SpaceX would have many options, such as the client buying (leasing?) a core outright and only paying the incremental costs of launch, recover, refurb, and rent (storage).
Plus, with the assembly line rate of sattelite production, Starlink could afford a few launch failures, so they could go deep into the F9 reuse count with continuing reductions in cost per launch.
Then further savings on refurbed soaked fairings sans acoustic material vs first flight articles.

Lots of cost savings a one off, multi-year satellite builder would not want to take.

 

[Singer3000]

You say owners as if SpaceX could not have a large stake in Starlink.

As to pricing, SpaceX would have many options, such as the client buying (leasing?) a core outright and only paying the incremental costs of launch, recover, refurb, and rent (storage).
Plus, with the assembly line rate of sattelite production, Starlink could afford a few launch failures, so they could go deep into the F9 reuse count with continuing reductions in cost per launch.
Then further savings on refurbed soaked fairings sans acoustic material vs first flight articles.

Lots of cost savings a one off, multi-year satellite builder would not want to take.

I strongly suspect SpaceX will retain 50%+ ownership of Starlink and treat it as a consolidated entity.

 

[Matias]

You say owners as if SpaceX could not have a large stake in Starlink.

As to pricing, SpaceX would have many options, such as the client buying (leasing?) a core outright and only paying the incremental costs of launch, recover, refurb, and rent (storage).
Plus, with the assembly line rate of sattelite production, Starlink could afford a few launch failures, so they could go deep into the F9 reuse count with continuing reductions in cost per launch.
Then further savings on refurbed soaked fairings sans acoustic material vs first flight articles.

Lots of cost savings a one off, multi-year satellite builder would not want to take.

I’m not disagreeing with your points. I just say, that after split up, Starlink is just a customer to SpaceX. What do you want from a customer? Money.

 

[Krugerrand]

Starship would certainly bring launch costs way down, but let’s not forget it has never flown. I have no clue what the actual costs would be or what SpaceX is paying for their satellite hardware. Designing, building, and flying the largest rocket ever will invariably encounter some difficulties and delays. I wouldn’t count on it for the initial constellation deployment.

Blue Origin will have their own rockets to keep costs down. Jeff Bezos also has a bottomless checkbook to get the company going.

I have no faith OneWeb will survive without cheap launches.

Our first Starlink carebear.