Birthright Margin

The “TPUs = bad for NVDA” take is up there with the dumbest yet (maybe worse than DeepSeek) as it *completely misses* what actually happened in just the last 6 weeks. And I will remember *who is who in the zoo*. My view… 1) Demand for AI is bananas (no one can meet demand, everyone is spending more). $GOOGL said just yesterday they have to double capacity every 6 months to keep up. 2) Scaling laws are intact (Gemini 3). The flywheel is about to speed up. Somehow the mid-curve crew thinks this is “zero sum competition”? NONE of this suggests that. 3) if you think the race is hot now, wait until you see what comes out of large coherent Blackwell clusters. All the magic from the “God Machines” is pretty much still Hopper based… Lastly, a quick GPU/TPU lesson: the cost and performance specs on the box aren’t what you get in real life. Duh. And $GOOGL is going to get a fat margin too. Double Duh. What matters is system-level **effective** tokens/watt/dollar and TCO. $NVDA GPUs have higher **MFU** because they’re already embedded in workflows / the ecosystem is massive. Btw, this is a good test… if you have an “opinion” on this topic but have to look up MFU, then perhaps curate better sources (BAD follows can be expensive!) The above effective token / watt gap also likely widens with Rubin CPX. Add in that Jensen can actually deliver volume into a tight market (you can get more of his chips sooner) + future flexibility (multi-cloud capable, programmable for paradigm shifts) and he’ll sell every GPU he makes for YEARS. $GOOGL will too since everyone wants a second supplier and TPU is a fantastic chip (I love $GOOGL). But this is about as far from either/or as it gets. The one benefit of this confusion is that it is likely to give $GOOGL a brief stint as ‘World Heavyweight Champion’ (most valuable). I would guess the midwits put the strap on them in less than 2 weeks.