Jensen just gave the clearest read on the neocloud financing question — and it's basically "demand solves everything."
His logic: you can't manufacture demand, but when it's real and high, capital finds a way. The neocloud bottleneck isn't customer interest or utilization — it's working capital to build out capacity. And working capital, he's saying, will show up because the underlying fundamentals (strong customers, real demand) are there.
This is the classic supply-constrained buildout story. If you're running hot on utilization and have enterprise customers lined up, the financing gap closes. Banks, private credit, strategic partners — they all move when the revenue visibility is clear.
What Jensen's signaling: the neocloud thesis isn't broken. It's just capital-starved in the short term. If demand holds (and he clearly thinks it will), the financing will materialize. That's a different risk profile than "no one wants the capacity."
For $NBIS $CRWV $IREN $CIFR $APLD — this is the kind of comment that separates temporary liquidity stress from structural demand problems. If you believe Jensen's read on AI infrastructure demand, the working capital gap is solvable. If you don't, the financing concerns are just the symptom of a bigger issue.
I'm inclined to trust the guy who sees every GPU order and every data center buildout plan before anyone else does.
His logic: you can't manufacture demand, but when it's real and high, capital finds a way. The neocloud bottleneck isn't customer interest or utilization — it's working capital to build out capacity. And working capital, he's saying, will show up because the underlying fundamentals (strong customers, real demand) are there.
This is the classic supply-constrained buildout story. If you're running hot on utilization and have enterprise customers lined up, the financing gap closes. Banks, private credit, strategic partners — they all move when the revenue visibility is clear.
What Jensen's signaling: the neocloud thesis isn't broken. It's just capital-starved in the short term. If demand holds (and he clearly thinks it will), the financing will materialize. That's a different risk profile than "no one wants the capacity."
For $NBIS $CRWV $IREN $CIFR $APLD — this is the kind of comment that separates temporary liquidity stress from structural demand problems. If you believe Jensen's read on AI infrastructure demand, the working capital gap is solvable. If you don't, the financing concerns are just the symptom of a bigger issue.
I'm inclined to trust the guy who sees every GPU order and every data center buildout plan before anyone else does.