Nvidia (NVDA) has cornered the market on artificial intelligence, according to Bernstein Research Managing Director and Senior Analyst Stacy Rasgon. When asked by Yahoo Finance Live about concerns there could be a bubble forming in AI, Rasgon says the fact that demand is so strong and supply is so tight, you could get worries about an air pocket."

Video Transcript

- So Nvidia is far and above the leader when it comes to AI. Do they kind of have the market basically cornered in this area?

STACY RASGON: Yeah, I mean, right now look, so there's only a couple of companies-- there's a lot of companies out there that have like an AI narrative about them. There's only actually two that are showing real actual dollars right now. It's Nvidia and Broadcom, and that's about it.

And Nvidia is like here, and then you've got everybody else, which is like pretty far below. So they do kind of have it quartered right now. There are other players that are trying to develop products for this market. You've got like, AMD there's a story as they get into next year and they have other products that are there.

And you've got Intel and you've got some others that are trying to do this. You also have some of their large hyperscale customers like the Google's and the Amazons of the world that are building their own chips, which is not new. They've been doing their own chips for many, many years. But you have that aspect as well. But in terms of benefiting from the swell of demand that has occurred with AI probably since the beginning of this year, it really is Nvidia that is in the driver's seat.

- And the hype that's surrounding Nvidia largely began around their previous earnings announcements where they just announced this guidance that blew past analyst expectations. And I believe your firm has done an analysis whereby their four biggest customers could easily deliver that guidance. And so anything else, presumably would be gravy.

STACY RASGON: Well, so you're right. So last quarter's guidance was pretty astonishing. I had to look at the number twice. I couldn't register in my brain, and I've been doing this job a fairly long time.

I've never seen a guide like that from an established like, at-scale company like we saw. So that was pretty unbelievable. If you do the math, so they're guiding total revenues, and again, I think, expectations are that they'll do better than this. But they're guiding total revenues at about $11 billion for this quarter.

If you work through it, the implied guidance for data center this quarter is about 8 billion, which is roughly a $4 billion incremental data center growth this quarter versus last quarter. If you sort of do the math on some of the larger hyperscale customers, and they've all guided CapEx, the CapEx increases from four of them would be enough to account for that $4 billion incremental. There are clearly lots of other customers besides those four that are buying these parts. But I think that the companies like at least implied guidance the way they laid it out for this quarter is easily doable given what we've seen in the supply chain and given their customer demand.

- So there's been a little bit of talk about this, not much yet because given the early days about the potential for a bubble, an AI bubble. I don't know that we would say that Nvidia is there. Is there the potential for that? And how long-- what's the kind of trajectory for that, if there is?

STACY RASGON: So at some point, yeah. So I mean, right now, demand is off the charts, but again supply is very limited. So customers can't get what they want. And there's a general behavior that happens in semiconductors and lots of other places, it's called double ordering. When people can't get what they want, they order more.

This is the toilet paper situation from like, 2020. It's the same kind of dynamic. And that sort of thing, like forget Nvidia for a minute, just in general, that sort of behavior happens. And so the fact that demand is so strong and supply is so tight right now, you could potentially worry about an air pocket or something.

And with Nvidia, that's one of the primary controversies is is this new level of demand that we see, is that sort of the new baseline for growth? Or is it like pull forward and panic buying and is there an air pocket? And look, at some point, what is the chance that Nvidia will see an air pocket?

It's probably 100%. We've seen them before. We'll see it again. I mean, look, the stock may have a 50% drawdown in it again at some point. We've seen those before.

It might be $1,000 to 500, like, I don't know. We'll see it. Am I worried about an air pocket like right now? Absolutely not, like, they're certainly not going to see one this year.

My guess is even next year is just fine, like, calendar 25, like, by the time we get there, maybe we'll know a little more. But like, I guess, we'll worry about that a little when we get a little closer. And frankly, the actual run rate that we could be seeing by the time we get there could be materially higher than what we see.

And I will say along those lines, if I'm looking at five years or 10 years, I'm convinced that we'd be talking numbers of that point that are materially higher than what we are talking about today. We are still very early in this whole thing, like, ChatGPT and all this other stuff that's gotten people excited, it's only been here since November. It's less than a year. We are very early on the whole progress of the AI cycle.

They'll probably overshoot. They'll probably undershoot. This is semiconductors. This is why we get cycles in the industry. I'm not worried about at least the next 12 to 18 months, maybe even longer, I think, we're OK.