Gil Luria

The race is still on.

Which means that OpenAI will spend the $250 billion on Microsoft, which now has by far the biggest backlog in AI compute.

And then Microsoft, Amazon, Google, and others will turn around and spend that on NVIDIA chips.

And so we have Nvidia and Microsoft with the best business they've ever had, for Microsoft maybe in 25 years, for Nvidia forever, they're both trading in the low 20s on earnings, which are historically low multiples.

So it's a historic opportunity in Nvidia and Microsoft precisely because OpenAI will stay in the race.

Everybody still needs NVIDIA.

All this AI compute will overwhelmingly be built on NVIDIA chips.

There will be some other chips because all these customers of NVIDIA want desperately to diversify, but they can't do it right now because they're building out their capacity so quickly.

NVIDIA has the best chip, has the capacity, it works, it's available.

Everybody else will start their own chips, but the only company that really has made any significant progress on their own chips is Google.

And that's because they started 10 years ago.

That's how long it takes to have a good chip.

Ask Amazon.

They're five years in, and their chip is just starting to get good.

Didn't look too bad in the earnings.

Microsoft, Meta, and OpenAI can start building their own chips, but they're very far away from those being any significant portion of their compute.

It's going to be overwhelmingly NVIDIA.

They exceeded very high expectations.

Their growth accelerated from more than 60% to more than 70%.

There's no other software company even close.