Post

Google has pushed Nvidia down to fourth place, and there’s more to it than meets the eye. It’s not about how much Nvidia has fallen; it’s that the market is starting to recalculate the value of AI: who’s worth more, the pick-and-shovel sellers or the people using the shovels? For the past two years, Nvidia has been the sharpest spear in the AI narrative. Data center revenue has climbed to more than 80% of its total, GPUs have been impossible to get, and the market has been willing to assign it a valuation that includes a “compute premium.” But now the other side of the ledger is coming into view: the more Google, Microsoft, and Amazon buy, the more they spend on capital expenditures, and the more real the depreciation burden becomes over the next few years. Take Alphabet: quarterly capex is already in the tens of billions of dollars, and its guidance keeps moving higher. That spending will ultimately hit the income statement; you can’t make it disappear just by saying “AI is a long-term play.”



Last month, I cut my Nvidia position in half. The reason was simple: I couldn’t sleep before earnings. Watching it swing up and down afterward, I’d be lying if I said I didn’t regret it. But I haven’t touched my Google position: the search ad money-printing machine is still running, the cloud business has finally stopped burning cash, and it would take competitors years to catch up on TPUs. The answer is in the details. Google isn’t just someone using the shovels; it makes its own. It uses TPUs for both training and inference in-house, and can also sell compute capacity through its cloud business. That vertical integration lowers unit costs. Search ads provide cash flow, the cloud business has gone from losses to profits, and although capex is high, at least some of it can be turned directly into revenue. By contrast, the valuation of a pure-play pick-and-shovel seller depends more on customers continuing to ramp up capex. If growth slows, its valuation could get cut faster than its earnings.

Big Tech earnings are all anyone’s talking about on Twitter lately, so let me pour some cold water on the discussion: don’t just look at who beat expectations; look at whose capex can still be sustained. The more Nvidia sells, the heavier the depreciation burden on Google, Microsoft, and Amazon. They’ll have to reckon with that sooner or later. For example, if a GPU server is depreciated over three to five years, the depreciation peak corresponding to this year’s buying spree may still be ahead. Cloud providers also can’t easily tap the brakes if they want to stay competitive in AI. So the market is starting to split: for the same AI investment, some see revenue while others see costs. Weekend trading volume is only 30% of what it is on weekdays, so even a little volume can send stocks higher. Don’t read too much into ranking changes at times like this. Rankings are an outcome, not a cause, and shifts in the pecking order under low liquidity deserve less weight.

I’m also keeping track of what would invalidate this view: if Google Cloud growth falls back below 20% next quarter, or if capex guidance contracts significantly, then this “third place” ranking is built on sand. I’ll keep that in mind and come back to check the numbers.
View Original
post-image
This page contains third-party content and does not constitute any advice, nor does it represent Gate's endorsement of such views. For details, please see disclaimer.
NVDANVDA-6.85%
MSFTMSFT-1.96%
AMZNAMZN-1.95%


Add a comment
Add a comment

Comment 2
Belalelbanna
good luck today my friend
Positive Demon
Waiting for an update 🍿