Article Directory
Generated Title: Palantir's AI Numbers Are Soaring. Is Nvidia Next?
Okay, let's dive into this Palantir-Nvidia connection. The headline is all sunshine and roses: Palantir's revenue is up, Nvidia's chips are in demand, AI is booming. But my experience tells me to always look under the hood.
Decoding Palantir's AI Surge
Palantir's Q3 report is definitely eye-catching. Revenue up 63% year-over-year to $1.18 billion (analysts were expecting $1.09 billion, so a solid beat). Adjusted EPS at $0.21, eclipsing the $0.17 estimate. The real kicker, though, is the U.S. commercial segment, home to their AI Platform (AIP). That segment's revenue jumped 121%. Customer roles also surged, up 65% as new users pile onto AIP.
But let's not get carried away. A 121% jump sounds huge, but the actual revenue is $397 million. Still impressive growth, but it's important to keep the scale in perspective. And that TCV (total contract value) soaring 342% to $1.3 billion? That's future revenue, not money in the bank today. Palantir's remaining performance obligation (RPO) – contractually obligated revenue that hasn't been booked as sales – surged 66% to $2.6 billion. Again, promising, but it's future potential, not current reality.
Now, here's the part of the report that I find genuinely interesting: the connection to Nvidia. Palantir's results are being touted as evidence that the AI revolution is still in its "early innings." Jensen Huang, Nvidia's CEO, mentioned a backlog for Blackwell and next-gen Rubin chips reaching "half a trillion dollars." That's a massive number, no question. The "Four Horsemen" (Amazon, Microsoft, Alphabet, Meta) are planning to drop $380 billion this year alone on capex, mostly on data centers to feed the AI beast. And they say that spending will increase in 2026.
The argument is that Palantir's growth shows AI adoption is moving "downstream," beyond just the tech giants. As more businesses jump on the AI bandwagon, they'll need Nvidia's chips to build the infrastructure. Nvidia, with its dominant 92% share of the data center GPU market, is positioned to profit. Since the AI boom began in early 2023, Nvidia's stock has soared 1,320%. The stock trades at a premium – 31 times next year's expected earnings. But the author asserts that it's a reasonable price to pay considering the projected 27% annual revenue growth over the next five years.
I've looked at hundreds of these filings, and I always ask myself: are we seeing real demand, or just a self-fulfilling prophecy driven by hype?
The $38 Billion Question
Then there's the Amazon-OpenAI deal. Amazon's stock jumped 4% after announcing a $38 billion deal with OpenAI for access to hundreds of thousands of Nvidia's AI chips via AWS. OpenAI will run its AI workloads on Nvidia's GPUs within Amazon's AI servers, scaling to full capacity by the end of 2026. Amazon stock jumps on $38 billion deal with OpenAI to use hundreds of thousands of Nvidia chips
But consider this: OpenAI also has a $300 billion deal with Oracle and agreements with CoreWeave worth over $22 billion. They've also announced deals with Broadcom, AMD, and, yes, Nvidia again.
These "circular AI deals" are starting to raise eyebrows. Are we in a bubble? Is AI demand being overstated? OpenAI's costs are projected to exceed $1 trillion by the end of the decade, while its revenue is far below that. Can they even afford these infrastructure commitments?
And let's not forget that Amazon's AWS is also providing OpenAI's rival, Anthropic, with 1 million of its custom AI chips by the end of 2025. So, Amazon's hedging its bets. Smart move.
Here's the methodological critique: How are these "demand" numbers being calculated? Are they based on actual usage, projected needs, or just aspirational targets? Are these "deals" binding contracts, or simply non-binding letters of intent? The devil's in the details, and those details are often missing.
Show Me the *Real* Demand
Palantir's numbers are impressive. But the AI landscape is still murky. It feels like we're in the middle of a gold rush, with everyone scrambling for picks and shovels (Nvidia's position), and cloud services (Amazon, Oracle, etc.) But are they selling to actual miners finding gold, or just selling to each other in a frenzy of speculation?
A House Built on Sand?
It's too early to tell if Nvidia's surge is sustainable. Palantir's growth is encouraging, but the circular nature of AI deals and the lack of transparency around actual demand raise serious questions. The promise of AI is undeniable, but the current valuations may be built on hype rather than solid foundations.