Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

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

cnbc.com

Post reply on HN
Print view View original

731–740 of 1001 posts

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

#731

Post by jsnell »

otterley wrote: post #700
jsnell wrote: post #687

Earlier quoted context omitted.

But the OP very clearly did not write anything of that sort. Their claim was: > This deal increases SpaceX's revenue by $11 billion per year. And that is pretty obviously correct. This deal is Google is buying a service from SpaceX for $920M/month, not investing in SpaceX. And that is revenue for SpaceX. I don't know why you're so insistent it isn't.

The false—or at least highly questionable and unsubstantiated—claim is “Now with this incredible deal, SpaceX is now GAAP profitable under the existing rules” simply because Google bought $11B of compute from SpaceX. It depends on how much it costs SpaceX to provision and operate the compute, and it depends on what other expenses and revenues they have. A quick peek at their S-1 filing shows a $5B annual loss last ye…

Oh, if that was your objection, why did you identify the issue as "But it is not revenue for SpaceX, which is the error OP made"?

> A quick peek at their S-1 filing shows a $5B annual loss last year. Unless SpaceX is selling compute to Google at a 50% margin (unlikely but possible), they’re not going to turn a profit because of this deal. Any profit that does result will be small.

The cost of AI data centers is almost entirely the capex (10% opex, 90% depreciation), so the costs aren't meaningfully affected by whether the DC is idle or operating at full load. They're renting their DCs to Anthropic and Google for a combined $25B/year. The loss of the AI division is about $2.5B/quarter. The math is pretty obvious.

> Google’s equity investment and P/E multipliers are irrelevant and have no bearing on SpaceX’s profitability.

Indeed. But the OP did not claim that either.

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

#732

Post by hijodelsol »

redox99 wrote: post #567
hijodelsol wrote: post #522

Earlier quoted context omitted.

That's roughly 11.66$ per hour per GPU, which is above the price that Google Cloud is reselling them at if you commit for 36 months: https://getdeploying.com/gpus/nvidia-b200

That's for standalone servers. xAI has something that nobody else has. The single largest interconnected cluster[1]. For inference it doesn't really matter, but for training that networking is crucial. [1] Probably, there could be undisclosed clusters owned by other companies.

I would think that for training Google has optimized their entire setup for TPUs so I can’t see how this would be used for anything but inference

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

#733

Post by WarmWash »

cavemandaveman wrote: post #636
sebastos wrote: post #563

Earlier quoted context omitted.

And you can sell your tulip. But if the mania stopped and you suddenly _couldn’t_ find another person to sell it to, would you now be upset you paid $5000 for a tulip? What’s the value at which you wouldn’t be upset? Ok, that’s the intrinsic value of a tulip to you.

Berkshire Hathaway doesn't pay a dividend yet the business has steadily grown more valuable

Because dividends are stupid and Berkshire is smart. Share buybacks are the optimal way to do "dividends".

The only reason to do a dividend is because people like the feels of getting a cash payout.

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

#734

Post by s1artibartfast »

manlymuppet wrote: post #633
tristanj wrote: post #205

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…

But Google loses $11 billion per year, and they gain $50 billion... in stock? As far as I know they really will be paying $11 billion annually in liquid cash to SpaceX (not a small ask) starting this year, and all they get in return is more money on paper? What incentive do they have to help SpaceX out like this at great cost, if they're not actually buying something valuable? Why are they incentivized to do this if…

They also get compute, which has real value. If you are going to spend 11B on new data centers or rental, better to spend with a a company you are about to ipo.

Either way, 500% return on the spend would be amazing

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

#735

Post by SecretDreams »

gretch wrote: post #662
Gareth321 wrote: post #580

Earlier quoted context omitted.

Bonds are no longer recommended. Current research indicates 100% equities to be the best composition leading up to, and past, retirement. To point, the economic uncertainties around geopolitics, AI, and war, plus irresponsible debt spending by governments and the prospect of QE (and higher inflation), is pushing long term rates steadily higher. There’s a reasonable chance that 30y treasuries are nearing 6% by the end…

> Anyone who holds bonds in this market will likely lose money. Yes, you lose money (or more precisely you lose opportunity) but you gain certainty. Which is what you want for retirement That’s pretty much the definition of risk premium.

I sleep on certainty. I feel bad for the people based their futures entirely on a trajectory from a time we'll look back on as "utterly unsustainable".

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

#736

Post by toomuchtodo »

SecretDreams wrote: post #735
gretch wrote: post #662

Earlier quoted context omitted.

> Anyone who holds bonds in this market will likely lose money. Yes, you lose money (or more precisely you lose opportunity) but you gain certainty. Which is what you want for retirement That’s pretty much the definition of risk premium.

I sleep on certainty. I feel bad for the people based their futures entirely on a trajectory from a time we'll look back on as "utterly unsustainable".

If you don’t have hope when you have little else, you don’t have anything. The behavior is understandable, even if wildly irrational.

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

#737

Post by fc417fc802 »

runako wrote: post #281

Since the S-1 filing, xAI has taken over and is likely the largest share of revenue. I would estimate that ~95%+ of xAI revenue, and 100% of its profit, is from renting their datacenters. This is a datacenter REIT bolted onto a social media company bolted onto launch business bolted onto a niche ISP. The expected price to sales is ~100x. The best datacenter REITs trade at ~10x and pay a dividend, which SpaceX does no…

> Comcast is one of the best-run ISPs

You mean the company with such a bad reputation that it had to aggressively rebrand? I take it you've never had the displeasure of doing business with them.

That said it wouldn't surprise me in the slightest to learn that it was one of the most profitable ISPs for investors. That would fit quite well with the general theme of prioritizing the interests of investors over all else.

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

#738

Post by toomuchtodo »

gordon_freeman wrote: post #603
rootusrootus wrote: post #336

Earlier quoted context omitted.

TSLA has a forward PE of ~200x. That is probably the most logical comparison with SpaceX. Proof that the market can stay irrational for quite a long time. It fills me with a bit of dread about the future of the market. I am 10 years out from retirement, have a bit over 1M sitting in that market, and I wonder if it will implode in the meantime. I am fairly committed to the "invest like a dead man" (i.e. index funds, n…

In a similar situation: I basically have just 2 funds in my retirement portfolio: SnP500 index fund (75%) AND Berkshire Hathaway B shares (25%) from my research I know that in years where SnP500 drops too much (recessionary periods), BRK-B would soften the blow as Value stocks tend to do well in such times. And usually that works for me.

What about swapping the SP500 for VT (total world equities)?

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

#739

Post by jnwatson »

missedthecue wrote: post #324
theturtletalks wrote: post #287

Earlier quoted context omitted.

Do companies like Uber, Tesla, etc ever intend to pay dividends? If a stock never intends to pay dividends, the value of the stock is simply the price the next shumck is willing to pay.

The value of the stock is your share in the underlying business. Because underlying business changes over time (hopefully for the better) you are not simply hoping another shmuck pays you more, like with tulips, whose underlying value does not change with time. You own a portion of a concern that is improving its own fortunes. Furthermore, dividends are approved by the board once per quarter or once per year. A divid…

Among the oldest value models, the Dividend Discount Model, says that the value of a company's stock is based on the present value of its future dividend payments.

Even if a company doesn't currently pay dividends, it will eventually do so or be purchased by a company that does. That's the theory at least.

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

#740

Post by notahacker »

JumpCrisscross wrote: post #685
adgjlsfhk1 wrote: post #673

Earlier quoted context omitted.

Even permitting isn't a clear win. You are changing from land permitting (where you can pick the location to be wherever you want) to launch permitting (where you have to coordinate with the federal government for airspace and water closures). Not to mention that with the current regulatory status, a rocket explosion can easily lead to a multi-month mandatory safety review that blocks all new launches.

> changing from land permitting (where you can pick the location to be wherever you want) to launch permitting (where you have to coordinate with the federal government for airspace and water closures) One of these is orders of magnitudes longer and more complicated than the other. Land permitting always involves multiple layers of government. And most of them are causing months- to yearslong delays. (Power hook-up i…

Plenty of world to jurisdiction shop beyond the US, most of which also has cheap land and plenty of sun, and for better and worse much of that world is less regulated than launch (and FCC spectrum licensing if you want your data back) and easier to skip the queue with comparatively small amounts of money. Hell, if you like your unit economics to be dependent on solving physics problems most of the earth's surface doesn't need permits at all...
Post reply on HN