Hacker News Reader

Comment by p-o | Hacker News Reader

Hard to not look at the crazy valuation of SpaceX and not see a correlation. At some point, something's gotta give.

Unfortunately, their valuation has almost nothing to do with starlink revenues. It’s almost entirely speculative oribital data centers that have not been invented yet. They could double their starlink revenues and it would have no impact on the valuation.

Might as well base the valuation on SpaceX getting a colony to Alpha Centauri and then billions living there and needing regular SpaceX shuttle services from Sol.

Then we could properly value the IPO in the quadrillions. (I know you know this is ridiculous. But investors clearly are riding high on the tulip mania).

In the meantime, until Musk comes up with the next big "idea" to switch to, all the current revenue levers need to be pushed to max to try to make it to that next stage, since public companies demand revenue every quarter.

A lot of their revenue is also from renting out GPUs to more productive AI companies, so depending on how long that game can keep on going (getting access to GPUs before other companies), it could last for a while too.

Here’s an idea: Musk launches a ton of orbital gpus but in reality it’s just an empty shell and real gpus sit in unused grok data centres. How would we even know?

That's actually the best take I've heard. It's like Amazon faking-it-till-they-didnt-make it with cashierless grocery stores.

Wasn't it Jeff Skilling who once waxed poetical about hypothetical future value?

I disagree, but not for the reason you think. They need to fund R&D to justify their high valuation. A new stock issue would decrease the valuation, and they will have a difficult time borrowing the money with their poor bond valuations. So this will (maybe) slow the bleed.

Musk doesn't need to deliver to keep valuation up. We've seen him doing pretty well stringing people along indefinitely with "3 months maybe, 6 months definitely" self driving capabilities.

It’s not an opinion it’s laid out clearly in the prospectus. Spacex is not a space company. It is a 2 trillion dollar AI company with a small launch business and a smaller ISP.

I am not disagreeing with that. I am saying that having enough revenue is necessary to keep the house of cards standing.

Even if invented, what is the advantage of orbit data centers?

Regulatory problems are doing the heavy lifting. Sure power is "free", if you can launch it, station keep it and cool it and if the data services are useful at the modem latencies of geosync and lunar orbits. But if datacenter projects can't be built/powered fast enough due to permitting, this is an expensive workaround.

The orbits are lower than that. Not that it changes the point.

Also, the lifetime of the panels on the ground is longer by about the same multiple as the power advantage of going to space, so if you're limited by the size of your photovoltaic factory, there's no reason to go to space.

The best I have heard: It makes you fully independent of the terrestrial energy grid. There is also some nebulous freedom from governmental authority possibilities.

Both of which strike me as poor arguments. We are in a short term energy crunch. I expect within five years that most of the energy problems will be resolved and the eye watering financials of launching a GPU into space will look absurd

Governments also do not like to cede authority, so I find it hard to believe that they will not claim jurisdiction over any business operating within their borders (all of those ground stations).

if you launch enough solar panels to power the data center into space, its easier to put them on the ground. Yeah there's some less power, but also have six times the lifespan so just build twice as many and it's still more efficient. Getting land in Texas can't be that hard.

You could build them at the South Pole, or at the bottom of the ocean and it will still be easier than putting them in space. Just think of maintenance.

Oh no, your power supply burned out Space: Welp, that entire data center node is dead now. Antarctica: OK, put a technician on a modified C-130.

I think the idea is easy energy availability and no push back from the locals. Those both seem to be to be fairly simple to solve on the ground compared to the unsolved issues doing it in orbit.

The Laws of Physics are much easier to build around than the Laws of Men.

No planning restrictions. Throw up as many as you can afford to build and launch. No grid strain because they're self-powered.

And they're likely out of national jurisdictions, so you can generate all the porn you want.

Technologically? Maaaaaybe a slight ping time reduction in certain configurations compared to terrestrial fiber, because even with multiple hops lasers can still be faster than fiber.

Super-expensive, questionably economic way to do it though.

> And they're likely out of national jurisdictions, so you can generate all the porn you want.

Means also no protection from foreign actors.

No national jurisdiction is a two-edged sword

Why cant we have self powered DC's on earth?

It is possible they delayed this rise as they didn't want bad news in the run-up to going public; but the is absolutely no way this can touch their fundamental numbers.

Reply on news.ycombinator.com