Yahoo Finance

Welcome to market domination. I'm Julie Hyman, that's Josh Lipton, and we were just listening to Nvidia CEO Jensen Wong making some announcements at Nvidia's annual GTC event, um, most recently showing off a robot there on stage. Joining us now is Cody Yak Acray, uh, Benchmark Company managing director and senior semiconductor research analyst. Cody, obviously, you were watching as well. Um, any big takeaways here? It was sort of flushing out some of the bones which

we were aware of already, right?

I think that's really been more, um, of the the nuts and bolts of the keynote was uh, really just a bit more details around what we knew was already in the works. Uh, Ultra and Rubin, uh, and now Fineman, uh, in 28, uh, that's that was incremental. We have not heard much about that in the past. Uh, but, uh, for the most part, it was a lot of incremental details that were relatively small.

Cody, you know, I I remember when Nvidia last reported financial results and we had, you know, other smart financial analysts on like yourself, your colleagues on the street, and I asked them, hey, okay, what is the next big catalyst for the stock? And some of them, Cody, you know, told us it was GTC, but I look right now the stock's down about, you know, nearly 2 and 1/2%. Was it was there something, Cody, investors were hoping for that they didn't hear today?

They were just hoping for more, um, maybe more details, maybe, maybe something about an acceleration of Rubin, uh, but with Rubin, it looks like, uh, it's not coming until the second half of of 26 now. Um, that, you know, that maybe the street had been more optimistic about its its launch. Uh, but, uh, I think the realism is, um, is that Blackwell Ultra is going to carry the, uh, the water here for the next 12 months, and then we'll be looking at, uh, at Rubin beyond that. Um, but it is, uh, obviously a bit of a disappointment from street expectations with the stock being off today.

Um, Cody, I I can't help but think of a parallel perhaps with Apple, right? Another company, obviously more consumer facing, but another company that has annual conferences where it rolls out new product, but maybe a little bit of diminishing returns after a while. In other words, obviously it was a huge leap when Nvidia first came out with Blackwell. Is it inevitably not going to be able to match the magnitude of that leap on a regular basis?

Yeah, yeah, Julie, I think you're right. I think that, uh, once we have the, uh, the the basic schematics or the basic roadmap down, that gets factored into the stock very quickly. And unfortunately, uh, the AI trade has been soft for, uh, the better part of this year, really since deep seek started to take the wind out of, uh, the AI trade, uh, a couple months ago. And and that has, uh, been looking, we've been looking for a reason to get back involved. And and the hope had been that GTC would be that reason, uh, but I think that was an overestimation of what, uh, Nvidia was going to say. I I really was not surprised that we didn't hear, uh, more specifics or, uh, more encouraging, um, stock catalyst out of this. I I don't think that's what this format was for.

So Cody, with the stock down now, 13% year to date, what gets it moving again in the right direction?

Well, it's just going to be earnings, um, and and I and I think Nvidia goes back to a block and tackle story, uh, that they continue to to deliver upside, they continue to guide higher, uh, they continue to grow into, uh, the transition which was delayed from Blackwell as they get into Blackwell Ultra, and that starts to get them, uh, more momentum to their earnings base and and uh, that's what investors are ultimately paying for. And so I think we just go back

to a beat and raise story around earnings.

Well, and and that I mean that is this that remains the story, right? To be clear, do you think that there is any demand slowdown, threat of a demand slowdown, any kind of issue there with what's actually happening with the existing product?

No, I don't think there is. I think if anything, things are accelerating. I think you're seeing that in the CAPEX budgets of, uh, the the the cloud service providers. Uh, those are only increasing and we are seeing, um, the compute density as we're moving to more reasoning models. And even Deep Seek, uh, showed that reasoning models with more compute, uh, perform better. And so there the, uh, the the air that was taken out of the AI trade around Deep Seek, um, was I think, uh, overexaggerated and, uh, and inflammatory largely, uh, based on, uh, the trend of reasoning that just drives compute density higher. Uh, and I think that has been just misunderstood by the industry is that as we move to these, uh, higher reasoning models, that they are going to consume whatever compute, uh, uh, Nvidia is able to provide.

Cody, you mentioned a point there I just want to come back to because I in the hyperscalers specifically, because I think for a lot of Nvidia investors, Cody, you know, they kind of keep it simple. They just think as long as the hyperscalers, as long as the big cloud giants, you know, keep spending, spending, spending to build out their AI infrastructure. Hey, that's good news for Nvidia. That's more GPUs. You know, how confident are you though, Cody, that, you know, you're going to keep seeing this really eye-popping CAPEX CAPEX growth, um, in not just this year, Cody, but 2026, 2027?

I don't think, Josh, there's any reason to doubt that we will need more compute horsepower in the future than we have today. Uh, that AI is becoming pervasive throughout all industries, whether it's industrial, automotive, uh, pharmaceutical, uh, mining, uh, any kind of delivery logistics. Uh, we're seeing it, it become pervasive throughout, uh, multiple verticals. And the more they use it, the more they need to rely on the cloud service providers for infrastructure. And so you're seeing companies like Oracle come on as a major supplier of resources. Uh, that's relatively recent development. Uh, but it's come on as we've seen enterprises adopt AI and need those resources on an outsource basis. And so I think that the cloud service providers will continue to spend to capitalize on that trend, which I don't see any reason that it stops unless the implementation of AI stops and and there's no reason that that slows as long as productivity increases are are available.

So would it stop or slow if indeed some of the some of the worst concerns about the economy right now come to pass? In other words, as you know, of course, there's a lot of concerns about tariffs, about the effect on the economy economy. We've seen some hits to consumer confidence. Do you have any concerns that that is going to affect the future pace of AI spending?

I I I do, Julie. I think the, um, the pace of spending is going to be very economically sensitive and there is, uh, big concerns in my mind and at our firm of a slowdown in, uh, in our GDP trends and and the global impact the tariffs are likely to have. And that is going to be a direct correlation to, uh, enterprise willingness to spend. Uh, that has to impact. There has to be a certain sensitivity, uh, to economic direction. And I think that that is the biggest risk to AI. I don't think it changes the long-term adoption. I think it simply, um, causes a pause as, uh, corporations start to try to figure out, uh, what the next permutation of our economy is going to look like and and if that is slower, then spending will slow.

Cody, appreciate that instant analysis and reaction. Thank you for joining us.