Live data from Hacker News

Google to pay SpaceX $920M a month for compute capacity at xAI data centers

cnbc.com

511–520 of 1001 posts

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#511
post #162

Google renting infra from xAI, I did not see that coming. My understanding of what computers are doing, what companies are doing and what governments are doing seems to be getting worse day by day.

Financial shenanigans aside, xAI seems to be buying hardware at breakneck speed. So why not? https://techcrunch.com/2026/05/20/musks-xai-is-being-sued-ov...

IIRC the large majority of their hardware (at least one tranche, they might have gotten more later) was Elon effectively stealing it from Tesla for xAI, saying “I’m personally doing Tesla a favor, since they can’t fully utilize it currently”, and is now renting that (stolen) compute to subsidize SpaceX.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#512
post #402

Earlier quoted context omitted.

Space data centers need years of time to design, build, and deploy, 5-10 at least, and that's after they solve their multiple very difficult or impossible problems. How will they cool them? There are just simple ideas like giant structures to radiate the heat away, but you say you need to put lots of mass in orbit? Like fsd, will take decades to figure things out.

They also need Starship at minimum, which is now a 10+ year old project still exploding regularly. Starship is at minimum a 2030 project at this point. And even producing the volume of chips needed for the type of growth space data centers would need to have to justify this would be another decade if construction started now on those fabs.

A minimum of 2030? That seems excessively pessimistic.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#513
post #199

Earlier quoted context omitted.

> Google renting infra from xAI, I did not see that coming. Actually that seems to be fairly logical? Hardware is what xAI has , and it's in great demand. So sell what makes you money. The real story here is that that xAI hardware is going to be running Gemini and not Grok. Which is to say: Grok basically failed as a frontier AI and they need to pivot to a business model which makes money. Obviously not everything Mu…

Could you please describe the other satellite launch services whose prices are competitive with SpaceX?

SpaceX's launch business is great, but does not justify the majority of the valuation.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#514
post #415

Earlier quoted context omitted.

Plus electricity, labor, and maintenance?

Yeah I was wondering about this too. It seems like way too much per GPU considering the purchase price of a B200 was around $40k last time I checked. So if we naively ignore the price of electricity and maintanence, it would only take 5 months before renting is a worse deal than buying outright.

It’s a supply and demand thing. Google would definitely be buying from nvidia and setting up themselves if nvidia had the capacity.

SpaceX/xAi/musk are currently in a good market for “happening to own 100k cards we have nothing to do with”, and are exercising that control as hard as they can.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#515
post #508

Earlier quoted context omitted.

That is one of the columns. The headline makes my point succinctly. Your paraphrase of the column misses the crucial point. A Nasdaq index fund doesn’t buy a company unless it is in the Nasdaq. Under the old rules SPCEX was ineligible for listing. Now Nasdaq index funds all have to buy. Index funds by nature do not selectively buy stocks, if the stock is in the index, they buy, that’s their mandate. That’s the game,…

[flagged]

In the context of my message it is very clear that they is SpaceX. This isn’t a secret. Nasdaq has said that they are changing the rules specifically for this listing.

It’s clear you aren’t interested in a good faith conversation. Thanks for the discourse either way.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#516
post #300

This is a masterful piece of financial engineering by Google and SpaceX. Google purchased 10% of SpaceX over a decade ago. After dilution they probably own around 5%. SpaceX is valued at a whopping 94x revenue. This deal increases SpaceX's revenue by $11 billion per year. If SpaceX maintains this revenue multiplier, then this single deal boosts SpaceX's valuation by 94 x 11 billion = $1 trillion dollars. Google owns…

I sincerely hope the market is not willing to value this sort of deal at a P/E ratio anywhere near 94. Off the top of my head, there is a very well established business involving buying expensive things and leasing them to the companies that intend to operate them so they can sell services: aircraft leasing. AER is the biggest player and they have a P/E ratio of, drumroll please, 6. And I expect that GPUs, despite cu…

> I sincerely hope the market is not willing to value this sort of deal at a P/E ratio anywhere near 94.

It will very likely be valued much, much higher. The SpaceX IPO is, in itself, a marvelous piece of financial engineering (requiring co-operation among multiple actors) which has been a long time in the works.

- Right out of the gate nearly all retail investment platforms have dramatically reduced requirements for purchasing an IPO, most notably Fidelity, which previously required $500,000 in your account to participate in an IPO reduced (on Friday) this amount to $2,000

- Retail investment, despite being quieter in the post-WSB era, is at all time highs.

- Reports are that the SpaceX IPO is already highly oversubscribed, meaning there are many more retail investors interested than there are shares available.

- SpaceX has a wildy low float of only ~4% which means price discovery will be much slower then normal, especially with aforementioned demand

- All of these retail platforms enforce some sort of "soft lock-in" whereby you're excluded from future IPOs if you sell your shares within 15-30 days. So if you want to get out you're not going to be able to participate in Anthropic/OpenAI IPOs in a few months.

- Coincidentally, most of the major indexes (thankfully excluding the S&P 500) have adjusted their rules to require only 15 days post-IPO before inclusion and have no profitability requirements. Many also adjusted the rules so that low float IPOs have their weight multiplied despite the low float.

- Many retirement accounts, in one way or another, are required to track these indexes and will be forced to buy these SpaceX shares at a very likely frenzied price and further drive the price up.

SpaceX will very likely open with far more retail demand than shares, the insiders (VCs, employees etc) will still be legally locked from selling, retail investors are penalized if they sell, and so the demand will be high and supply very low.

If they can keep this demand hyped for just 3 weeks, price will still be elevated when retirement accounts are forced to buy... roughly the same time retail investor start seeing the penalty for selling expiring (meaning it is not irrational at all to be in the IPO, but it is irrational to sell before being listed in an index).

Fun fact: the other fascinating thing about this IPO is the terms for insider lock-in. At first earnings (Jun 30) inside investors unlock and can therefor liquidate 20% of their shares... but if the stock performs well, they can unlock and additional 10%. There are additional rules for continued unlocking of more shares depending on performance as time goes on. So everyone on the inside has a very vested interest in a spike in stock prices: not only will their stocks be worth more, but they can realize that value faster.

I would be surprised if SpaceX price doesn't explode in the first few weeks because for everyone involved this would make sense. It's only in August that we'll start seeing the really interesting things start happening.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#517
post #199

Earlier quoted context omitted.

> Google renting infra from xAI, I did not see that coming. Actually that seems to be fairly logical? Hardware is what xAI has , and it's in great demand. So sell what makes you money. The real story here is that that xAI hardware is going to be running Gemini and not Grok. Which is to say: Grok basically failed as a frontier AI and they need to pivot to a business model which makes money. Obviously not everything Mu…

You make it sound like they’ve given up on Grok, which I don’t think is accurate. I think it’s been mentioned the Grok 5 1.5T model is currently training on Colossus 2. And their recent deal with cursor is part of being able to eventually compete with Anthropic for agentic coding.

There is no evidence they can deliver anything competitive based on their past performance, though.

Even if their model is competitive or even surpasses e.g. Deepseek (which is far from given) how would that justify a huge valuation?

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#518
post #508

Earlier quoted context omitted.

[flagged]

In the context of my message it is very clear that they is SpaceX. This isn’t a secret. Nasdaq has said that they are changing the rules specifically for this listing. It’s clear you aren’t interested in a good faith conversation. Thanks for the discourse either way.

[flagged]

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#519

“If we fail to deliver access to the committed amount of GPUs by September 30, 2026, then following a one-month grace period, Google may immediately terminate the agreement or accept the number of GPUs provided, with a corresponding pro rata reduction in the monthly fees. After December 31, 2026, the agreement may be terminated by either party upon 90 days' notice.” https://www.sec.gov/Archives/edgar/data/1181412/000…

The same terms Anthropic have today I believe.

Re: Google to pay SpaceX $920M a month for compute capacity at xAI data centers

#520

Earlier quoted context omitted.

According to their IPO S-1 draft they are 93% an AI company and 4% a space company. Its the remaining 3% of the company that is profitable, the Starlink stuff.

As I recall isn't Starlink revenue at least 3x Space revenue, so not sure how they are characterizing that 3:1 ratio as 3% vs 4% ! The "93% AI company" is also a huge mischaracterization since this isn't AI business - it's datacenter/GPU leasing business which their 2 customers can pull the plug on with 90 days notice.

Yeah, Starlink is about 3.5x the space launch revenue but still only about 0.5x in terms of profit. Falcon 9 is as optimized as a rocket could be, and absolutely owns the market. Starlink is a mostly rural service with global consumer pricing where average monthly rates in poorer countries drag the average down. Starlink government and commercial business, however, is growing quickly and I expect that soon Starlink will be ahead of launch, in terms of income, probably by the end of this year.
Post reply on HN