I'm not sure if you're refuting what I said, but there is absolutely no way you will get 20ms end-to-end latency from this. Satellite propagation delay, yes, but for your ping to get to Google and back will be over 50ms.

Why isn't space the biggest part of latency? can't they put a satellite link directly in google?

They have no cross-links, so up and down to space is going to happen multiple times in many cases. Second, they will not have data centers or ixps right where Google's ingress is, so it has to traverse fiber for quite a while. Someone gave some real numbers on the last SpaceX thread, but I believe that part alone will be 20-30ms added.

So how long until Google puts a datacenter in space?

Seeing as heat dissipation is a hard problem in space, probably never.

They do have cross-links though?

They do not have cross-links. They were in the original public announcements, but none of the current satellites have them, nor have they announced when they would be launching some that do. It's likely years away.

> Early satellites are launched without laser links, in October 2019 SpaceX expected satellites with these links to be ready by the end of 2020.

from wikipedia

That's a claim musk made, but they already have over 500 launched without them. They should have 1000 by year-end, so it's unlikely they'll be able to use them for a couple of years.

Would satellite on-board caching be possible? Could you cache, for instance, the most frequently requested cat photos currently on reddit? You would be able to bypass half of the round trip

Possible, yes, practical, no. In general, you want the least complexity as possible in the payload, since you can't fix it as easily as you can with ground bugs. Also, memory that's space-hardened is going to be significantly more expensive and failure-prone, so you need a lot of redundancy.

AFAIK Verizon is doing something similar with edge compute in their 5G towers/data centers. If Starlink gets big enough, I'd expect them to do the same. However, I think it's much more likely for Starlink to put the edge compute/cache at the Starlink ground stations, rather than in the satellites themselves. I think the tech/cost still isn't good enough to have significant compute on a satellite.

https://www.fiercewireless.com/5g/verizon-partners-aws-to-br...

I played CS 1.6 at 120-200 ping... people will manage.

I agree. Latency is much less of an issue than most people think. It's mostly about bandwidth.

It should be lower latency for transaltantic and other long haul traffic, as the speed of light in glass is roughly half that of the vacuum of space. So the extra distance to space is offset by the faster speed of light.

That is very roughly... it is about 2/3 c and can be calculated by the refractive index of quartz glass vs air (e.g. SiO2 has refractive index between 1.55 bad and 1.4 good for the speed) e.g. 1/1.55 ~ 0.645 and 1/1.4 ~ 0.714. According to "Main Parameters" section of [0] it seems to be 1/1.44 for the silica used in actual fibers.

A good approximation would be 200 km/ 1 ms instead of 300 km/ 1 ms for speed of light in a vacuum/ microwave links.

[0] https://www.rp-photonics.com/fibers.html

They don't have the laser interconnects working yet, so long haul traffic won't go over Starlink for some time.

Simulations have shown that it can still be faster than fiber even without laser interconnects.

That simulation only takes into account the satellite delay. Fiber delay will be another 20-30ms.