4 ms·

Consider: - SpaceX, the company building the network, is the same company who can deliver the payloads. - SpaceX is the only company who has re-usable rockets

• mikhailt 9y ago
  Everything they do has an impact on their end goal/master plan, Mars. They need to send thousands of ships to colonize Mars. This global sat internet service will require them to deploy tens of thousands of micro-sats to provide global internet. The revenue generated by said service will provide them the funding to build the said ships for Mars. Oh and of course, the eventual solar system internet.
  • gfodor 9y ago
    Yeah one can definitely imagine a future where SpaceX kills their payload-to-orbit delivery business in order to streamline operations and free up resources to focus on a) revenue generation from the sat network and b) the "get your ass to Mars" program. It'd be a risk from a diversification perspective but I could see them doing it once the sat network hits escape velocity wrt projected revenues landing them to cash flow positivity in the Mars program.
    • grey-area 9y ago
      They won’t have to. The larger rockets planned for mars (bfr) will make payload to orbit even cheaper as it is fully reusable, so only fuel costs.
  • pjc50 9y ago
    Colonizing Mars must not happen until we have a definitive answer as to whether there is or was native life there. The answer "there was life until Musk landed" is not going to be acceptable.
    • thaumaturgy 9y ago
      There will be no practical way to determine that until we've colonized it.
      • jandrese 9y ago
        Proving the existence of life is relatively easy: we just have to find some. Proving that Mars is lifeless is flat impossible. We can't even be sure our definition of life is sufficient.
  • Dylan16807 9y ago
    > solar system internet What exactly are you picturing here? Because of light-speed considerations you'd have semi-isolated networks at each major body, and bodies connected with special high-power high-latency links. And those links might as well be direct, only needing a handful of dishes. Is there a reason to scatter little relay satellites all over the middle of nowhere? Or do anything else that needs many many rocket launches?
    • stcredzero 9y ago
      What exactly are you picturing here? Because of light-speed considerations you'd have semi-isolated networks at each major body, and bodies connected with special high-power high-latency links. Late 21st century FidoNet. Late 21st century USENET?
      • phkahler 9y ago
        And mirroring of sites like Wikipedia, YouTube, and social media. Interplanetary email (with video attachments too).
    • toomuchtodo 9y ago
      http://ipnsig.org/ http://ipnsig.org/
• dkoubsky 9y ago
  Not to mention the competitor will most likely have to do without the subsidies SpaceX received from the government.
  • jandrese 9y ago
    Unlikely, all rocket companies get government assistance, usually in the form of launch contracts. SpaceX has launched several USGC payloads already for example.
  • greglindahl 9y ago
    You mean the competitive contract that SpaceX won to deliver supplies to the ISS? I've seen a few people claim that was a subsidy, but that's a bit far-fetched.
• dmoy 9y ago
  How much telecom traffic goes via satellite vs. ground towers and fiberoptic cable? What percentage of capex cost for a telecom is rocket launches? I have absolutely no clue, but I feel it would affect this analysis greatly.
  • gfodor 9y ago
    I don't know either, but considering the complete lack of existence of the kind of network SpaceX is proposing, or anything close to it, if it's not deployment costs it would have to be some other factor coupled with operating a ~1000 sat network.
    • pjc50 9y ago
      It's similar to, although much larger than, Iridium. Building it is one thing but making it profitable is another, as it making people accept the latency tradeoffs of satellite.
      • Game_Ender 9y ago
        There network will have minimal (30-40ms) latency hit vs ground based systems because there satellites are 300-1000 miles up vs the 20k miles that the current geostationary ones are at.
      • randall 9y ago
        Ars says the latency is 25ms. https://arstechnica.com/information-technology/2018/02/spacexs-satellite-broadband-nears-fcc-approval-and-first-test-launch/ https://arstechnica.com/information-technology/2018/02/space... They say it's due to lower orbits. Iridium is fewer satellites so they have to be higher. More w/ lower orbits means lower latency. Do you know if this is too good to be true?
      • smoyer 9y ago
        Most terrestrial systems are now using fiber optics ... the speed of photons in a fiber is about 2/3rds the speed of light. The transmission of radio signals in the vacuum of space is theoretically at the speed of light so: 40000 / pi / 2 = 6366 40000 / 2 * 3 = 60000 60000 / pi / 2 = 9550 9550 - 6366 = 3184 Per the above, any satellite Internet system below about 3000 km above the earth will have latency that's competitive with fiber optics on earth (note that I haven't done the simultaneous solution to account for the uplink/downlink distance so it's actually a bit less). For the SpaceX system, there should be better latency than the terrestrial Internet. Please ignore the rounding ... this is a back-of-the-envelope exercise. Fiber optics are great and far less susceptible to interference than radio - but if you Google articles on how high-frequency trading companies are installing line-of-sight microwave for their networks you'll see that fiber isn't our fastest technology.
• nordsieck 9y ago
  "Talk about an economic moat -- to compete on margin you'd first have to become a rocket company. And not just a rocket company but one who has technology to re-use rockets, which likely is a ways away from becoming replicated, nevermind commoditized to the point where it evaporates from costs." It's even worse than that. SpaceX is using micro satellites for its network; they can launch them for "free" (just the cost of extra fuel) as a secondary payload on any of their launches where there is excess cargo capacity.
  • tankenmate 9y ago
    They'll need to predominantly do them on polar missions, in 2017 SpaceX had only 5 polar missions. Also, for missions that are already launching large numbers of satellites (like Iridium flights; 4 of the 5 polar missions in 2017 were for Iridium) this may be harder as it will require more fuel and a relight after an extended mission in its primary payload orbit. EDIT: grammar fix
    • nordsieck 9y ago
      I don't think they're going to launch anywhere close to even 1/10 their network this way; there are too few scheduled launches and too many satellites in their planned network, even if they were able to overcome the technical obstacles you mention. Depending on the margins they earn, the few (if any) they launch this way may be a meaningful cost savings.
  • jagger27 9y ago
    But in the specific case of launch padding, another microsat company could just as easily buy that same extra capacity from SpaceX. Sure, that's assuming SpaceX doesn't cram every launch to the max. There's also the opportunity to cross-shop a bit with other launch systems.
• th1nkdifferent 9y ago
  No. Any company can use existing satellite launch companies to launch their constellation. You don't need to be a "rocket company" first. There are plenty that can do this on the cheap - ISRO for example.
  • gfodor 9y ago
    My argument is about margin and replication costs not about if there are alternative routes towards replicating SpaceX's network, given enough money. Of course if you had infinite money you could burn up hundreds of rockets to get your network in place using existing providers. But my guess is that money is going to flow literally anywhere else if, for example, you will be spending 10x money as SpaceX just for table stakes of getting your satellites in orbit.
  • phkahler 9y ago
    Nobody can do it as cheap as a used rocket. Since the they will be (relatively) mass producing their internet satellites, losing one on a highly reused rocket won't be much of a financial hit. SpaceX could test rockets to failure with their own sats for a tiny fraction of the cost anyone else would face.
• kilroy123 9y ago
  I think we'll soon have a better idea of how this plan might pan out. OneWeb is trying to do this first, and the first of many satellites are already built[1][2]. They'll start launching later this year. [3] I'm very curious to see a real-world use of their system when it's fully up and running. [1] http://www.oneweb.world/ http://www.oneweb.world/ [2] https://twitter.com/SpaceTechAsia/status/959300225394339841 https://twitter.com/SpaceTechAsia/status/959300225394339841 [3] https://spaceflightnow.com/2018/01/22/testbed-for-canadian-data-relay-network-successfully-launched-from-china https://spaceflightnow.com/2018/01/22/testbed-for-canadian-d...
• stcredzero 9y ago
  It's hard to argue that from an economics perspective this isn't an insane leapfrog type of event, assuming they can gather the necessary in-house chops to deliver good telecom service. Talk about an economic moat -- to compete on margin you'd first have to become a rocket company. And not just a rocket company but one who has technology to re-use rockets, which likely is a ways away from becoming replicated, nevermind commoditized to the point where it evaporates from costs. It's even steeper than that. SpaceX could arrange to send up some payloads piggybacked on their commercial payloads. Their competitors could effectively be paying to put part of their fleet up.
  • phkahler 9y ago
    >> It's even steeper than that. SpaceX could arrange to send up some payloads piggybacked on their commercial payloads. Their competitors could effectively be paying to put part of their fleet up. A friend of mine in the telecom business claimed that's how Level3 grew. They were laying fiber for others, but the largest cost of that is in logistics, permitting, digging, etc... So they just laid down their own fiber along with their customers at a fraction of the cost.