Asit Sharma

It is, I think, investing in its ecosystem.

I'm more a believer in this

Jensen Huang used this term AI factories way back when ChatGPT first exploded onto the scene.

He had a very clear vision.

He thought that these AI factories would have to be replenished every five years.

They would have to be equipped with the latest technology, not just GPUs, but networking equipment.

all types of storage.

And so, looking over the press release, look, part of this is that Core Eve has to adopt the Rubin platform, Vera CPUs, Bluefield memory.

So, NVIDIA needs proof points for other hyperscalers, for sovereign governments, for academic research institutions, that its AI factories are the one-stop shop for AI

That has to happen over the next five years.

This is why it's important for Nvidia.

Otherwise, if that wasn't in the offing, I would agree with you, Emily, that maybe they're over-investing here.

I do believe, though,

to circle around to your question, there is a point where Nvidia could become too diffuse in its investments.

We got to keep our eye on that because a couple of billion here could turn into 10 billion to 20 billion there.

Then you start getting into a true question of, is this all circular?

Right now, I believe that investors misunderstand the scale that Nvidia operates on, but we should watch the numbers.

If they start to mushroom,

Yes, it could be at the beginning of a circular type of revenue demand cycle, and that won't be good for anybody.

It's also Jensen's biggest fear, right?