> Starlink isn't designed to offer lower latency or lower prices than existing terrestrial wired or wireless ISPs (at least, decent ones). Instead it's designed to offer a low latency wireless service in places where that has never been available at all. Which is most of the land area of the Earth, and all of the ocean.
It is widely believed that a huge chunk of Starlink’s revenue, if not ~all of it, will come from leasing faster-than-currently-available low latency links to HFT companies for nyc-lon and such. This is why their first orbital planes being launched are for the north atlantic and CONUS.
Providing internet to the underserved areas is secondary; the cash will come from beating terrestrial fiber links (even idealized great circle ones) between markets that are already extremely well connected.
One such private low latency link between Seacaucus and LSEG will probably pay 2-5x what everyone else on the planet combined pays for internet access in underserved areas.
> It is widely believed that a huge chunk of Starlink’s revenue, if not ~all of it, will come from leasing faster-than-currently-available low latency links to HFT companies for nyc-lon and such.
This is speculation. Without inter-satellite links the latency between New York and London on Starlink would be quite disappointing.
> This is why their first orbital planes being launched are for the north atlantic and CONUS.
A LEO constellation can only be targeted at specific latitudes (not longitudes). They're just targeting mid-latitudes because that's where potential customers live.
The data on this map makes it look like they're going for pretty uniform coverage already. Is there a source for the early orbital planes having a specific purpose? The earth rotates underneath each plane, and the times of day it will be available to any given ground path change over the course of a year.
It does seem pretty convenient that the inclination they've chosen lines the satellites up nearly in a straight path from New York to London, but I imagine this would be relevant to many transatlantic communications.
There is a purpose, but it's not HFT... It's orbital mechanics.
You can't have efficient coverage of the whole planet without launching all of your satellites at fairly high inclination (>45°). There will be more satellites per area near the poles and so that's why Starlink will be available at very high (and very low) latitude before it's available at the equator.
> It is widely believed that a huge chunk of Starlink’s revenue, if not ~all of it, will come from leasing faster-than-currently-available low latency links to HFT companies for nyc-lon and such. This is why their first orbital planes being launched are for the north atlantic and CONUS.
True? I'd heard that the initial target market was passenger planes over the ocean, cruise ships, and similar maritime and aerial activity that has to use geosynchronous satellites at present.
Covid-19 has thrown a spanner into the works of that plan, for a while. I'm sure people with money won't be turned away.
Longer term, remembering that Musk hails from South Africa, I thought the plan was to bring internet to failed states or barely functional areas in Africa (outback Namibia/Zimbabwe/Tanzania etc) and possibly also the Americas and the Middle East).
The long term plan is to sell service to anyone who will buy it, to fund Starship for Mars. The ship and plane market can't be served by the current constellation because it doesn't have links between satellites. This also excludes the high frequency trading market (proving that SpaceX is not relying on traders to make their revenue goals). The initial market is rural areas of the US, and then other countries as they expand their ground station footprint.
Oh, sure, the latency will be lower - it's just that without these satellites it was simply not possible to get continuous coverage on long over ocean flights, so something is better than nothing.
While for airplanes I think the latency might not be that problematic (I think VoIP in plane package full of people would not be very welcome) you indeed want low latency connection on a cruise ship.
NYC-LON are already connected by shortwave systems built using huge-ass yagi-uda antennas and special bespoke modem + amplifier systems. Shortwave based systems have even obsoleted well known HFT microwave paths such as Chicago-NYC.
source: I know some of the people who built several of the Chicago-NYC links.
Be careful, everytime I make this point I get berated by the Spacex fan boys. The reality is, making high speed trading faster adds no true value to society. And therefore I find this entire project sad.
SpaceX is launching the initial constellation without the laser links required for the hypothetical low latency service for traders. This would never happen if traders were the only or primary market. SpaceX has stated that their target is the consumer market and their actions are in line with that.
You don't need inter-satellite links for low latency, as Mark Handley has elegantly pointed out in his incredible video. I don't think internet access will amount to even 50% of the revenue of Starlink.
You don't need satellites at all to beat fiber. Terrestrial microwave does it too. Starlink can't beat microwave except for places where it's impossible to install microwave, like over the ocean. But without cross links Starlink won't work over the ocean either.
Source? There is no indication whatsoever that the service will be cheap enough to be for the consumer market. In fact, all serious cost analyses show it's not.
It is widely believed that a huge chunk of Starlink’s revenue, if not ~all of it, will come from leasing faster-than-currently-available low latency links to HFT companies for nyc-lon and such. This is why their first orbital planes being launched are for the north atlantic and CONUS.
Providing internet to the underserved areas is secondary; the cash will come from beating terrestrial fiber links (even idealized great circle ones) between markets that are already extremely well connected.
One such private low latency link between Seacaucus and LSEG will probably pay 2-5x what everyone else on the planet combined pays for internet access in underserved areas.