CFRA Research's Angelo Zino Talks Nvidia's Revenue Outlook

Angelo Zino, SVP and Tech Lead at CFRA Research talks about Nvidia's blowout revenue forecast, the AI tech trade and more. He speaks with hosts Paul Sweeney and Alexis Christoforous.

See omnystudio.com/listener for privacy information.

Speaker 1: Bloomberg Audio Studios.

Speaker 2: Podcasts.

Speaker 1: Radio.

Speaker 2: News. The news of the day yesterday in terms of

Speaker 2: company earnings was NVIDIA. Some numbers after the close. The

Speaker 2: street likes it. The stock's up 70%. The number that

Speaker 2: jumped out at me was they projected revenue for the

Speaker 2: next fiscal year. Growth rate of 70%, 7-0. Now put

Speaker 2: that into context.

Speaker 3: It is stunning.

Speaker 2: And this is on a run rate of $ 300 billion.

Speaker 2: So it's not like they got a couple million dollars.

Speaker 2: This is $ 300 billion of revenue growing at 70%. So

Speaker 2: just extraordinary. And I think that tech space, tech investors

Speaker 2: breathe a sigh of relief that the AI investment momentum continues.

Speaker 2: Angelo Zeno, he does this stuff for a living. He's

Speaker 2: a senior vice president and tech lead for CFRA Research Corporation. Angelo,

Speaker 2: what were some of your takeaways from NVIDIA's earnings and

Speaker 2: conference call last night?

Speaker 1: Yeah, so I'd say the biggest takeaway, obviously, is kind

Speaker 1: of the revenue numbers that you guys had just alluded

Speaker 1: to in terms of expectations for 70% growth next year

Speaker 1: from mid-40s. And I think what's happening right now is

Speaker 1: a number of factors. One, hyperscalers are planning to spend

Speaker 1: much more than maybe we previously anticipated. I think Nvidia

Speaker 1: pointed to about $ 1. 3 trillion next year and that was

Speaker 1: about $ 200 billion above our expectations or forecast for next year.

Speaker 1: So I think that's kind of a big driver and

Speaker 1: also what you're actually seeing away from the hyperscalers, those

Speaker 1: Neoclouds plus Musk is kind of the bucket I would

Speaker 1: throw it in and that's actually growing faster than the

Speaker 1: hyperscalers and I think it's helping to really kind of

Speaker 1: diversify Nvidia's broader business. And then You know, what I

Speaker 1: would also say is the content growth story, and that's

Speaker 1: the biggest reason why we have been bullish here over

Speaker 1: the last couple of years, continues to exceed our expectations.

Speaker 1: And we see no signs of that letting up as

Speaker 1: we kind of look here over the next couple of iterations.

Speaker 1: And then, of course, I mean, I'd say the one

Speaker 1: blemish was the margin growth. um issue out there but

Speaker 1: that being said they are going to increase pricing as

Speaker 1: we go into um you know calendar 2027 and that

Speaker 1: you know clearly is supporting that 70 growth trajectory and

Speaker 1: helping to kind of maybe ease some of those pressures

Speaker 1: and stabilize the margins going into next year.

Speaker 3: Angela, let's talk about some of the possible hurdles for

Speaker 3: NVIDIA and for growth going forward. Are they too dependent

Speaker 3: on a small group of tech giants like the hyperscalers

Speaker 3: as customers? Also, remember, some of these customers are increasingly

Speaker 3: developing in-house chips that could potentially take business away from NVIDIA.

Speaker 1: Yeah, no, absolutely. You're absolutely right. And I'd say the

Speaker 1: competitive pressures are out there. And, you know, we still

Speaker 1: look at NVIDIA even after the monster move today. You know,

Speaker 1: I would say that the multiple remains compressed based on

Speaker 1: some of those concerns, some of the competitive pressures out there,

Speaker 1: maybe some of the circular financing concerns that are out there.

Speaker 1: as well. But as far as the competitive dynamics and

Speaker 1: the customer base is concerned, the good thing is we

Speaker 1: are starting to see, and there are signs that that

Speaker 1: customer base will diversify going into next year, essentially, NVIDIA

Speaker 1: pointing towards more of a 50-50 split as we go

Speaker 1: into next year. And that kind of helps alleviate maybe

Speaker 1: some of the concerns that from a customer concentration perspective,

Speaker 1: but that will remain for the foreseeable future. But in

Speaker 1: the same respect, listen, when you think about NVIDIA's ecosystem

Speaker 1: supporting about 6 million developers or so, those hyperscalers will

Speaker 1: need to continue to spend. I will just say this,

Speaker 1: maybe the growth rate from the hyperscalers going into next year,

Speaker 1: let's call it closer to 50%, whereas maybe for some

Speaker 1: of the non-hyperscalers, those neoclouds, you'll see significantly higher growth.

Speaker 1: Let's call it closer to 90% to 100%.

Speaker 2: Angel, one of the concerns out there for NVIDIA specifically,

Speaker 2: but I think just for the AI space in general,

Speaker 2: is some of the circular funding structures that are being

Speaker 2: created out there where maybe NVIDIA is funding some of

Speaker 2: its customers in various formats. What does the company say

Speaker 2: about that? Did they address that last night?

Speaker 1: They did. You know, they actually, I think for them,

Speaker 1: it's an area of strength and in many respects it is, right?

Speaker 1: I mean, historically, when you kind of look at the

Speaker 1: competitive advantage that NVIDIA has had, it has been that

Speaker 1: CUDA software, right? And now it's not just CUDA, it's

Speaker 1: also the fact that they've got some massive scale They're

Speaker 1: generating massive free cash flow. You're looking now at a

Speaker 1: $ 200 billion run rate. This time next year, it'll probably

Speaker 1: be closer to a $ 300 billion free cash flow run rate.

Speaker 1: And they're pumping a lot of that right back into

Speaker 1: the ecosystem. Essentially, what they're trying to do is build out,

Speaker 1: be a supporter in growing out that AI infrastructure, because

Speaker 1: without them, it's going to be nearly impossible to do

Speaker 1: For them to kind of go out there and support

Speaker 1: the ecosystem, they're helping to kind of lower the cost

Speaker 1: of capital, and they're actually allowing these Neo clouds to succeed.

Speaker 1: So in many respects, they're going to participate in terms,

Speaker 1: they're going to benefit in the sense that they're going

Speaker 1: to get that hardware revenue. They're also going to potentially

Speaker 1: get revenue in terms of and enjoy the upside of

Speaker 1: from the actual sales within those data centers. But it

Speaker 1: is a risk. It is absolutely a risk if we

Speaker 1: get to a point where we overbuild and NVIDIA is

Speaker 1: sitting there with, you know, having built too much capacity

Speaker 1: as well. So there are pluses and minuses to all this.

Speaker 3: Angelo, how much is NVIDIA doing all of this and

Speaker 3: just wowing us with their numbers without China? Because there

Speaker 3: still are restrictions on NVIDIA doing business there, right? How

Speaker 3: important is gaining more access to the Chinese market for

Speaker 3: NVIDIA's growth?

Speaker 1: Yeah, I mean, you would have thought it would have

Speaker 1: hurt them by now. I mean, to be honest with you, yeah, right.

Speaker 1: It was, I think, 1% of their revenue in terms

Speaker 1: on the data center side or less than 1%. So

Speaker 1: virtually no revenue coming from China at this point in

Speaker 1: time is all upside. But at some point in time, listen,

Speaker 1: it's going to be an issue for them as the

Speaker 1: growth rates here in the U.S. eventually have to slow down.

Speaker 1: And we'll see how successful, you know, we kind of

Speaker 1: get this open source versus closed model debate. But as

Speaker 1: you kind of see more and more token growth and

Speaker 1: success from the Chinese market, I think you're going to

Speaker 1: see greater growth. you know, maybe greater concern or motivation

Speaker 1: from NVIDIA to try to get the U.S. government to

Speaker 1: get some of their, you know, more of their chips

Speaker 1: into China. But at this point in time, they're going

Speaker 1: to have to do it without China because there's no way,

Speaker 1: you know, we think the U.S. government is going to

Speaker 1: allow that to happen.

Speaker 2: Angelo, from a competitive perspective, what is the biggest threat

Speaker 2: or challenge to NVIDIA? Is it the other chip makers,

Speaker 2: the Broadcoms, the AMDs of the world, or is it

Speaker 2: some of their customers as Alexis was suggesting earlier, that

Speaker 2: may be developing their own chips internally.

Speaker 1: I think it has to be some of their customers

Speaker 1: at this point in time in the sense that they're

Speaker 1: going to have to continue to purchase NVIDIA's servers. That's

Speaker 1: not going to stop. But what it can do in

Speaker 1: terms of you know, some of these hyperscalers developing more

Speaker 1: of their own internal chips is it will slow eventually

Speaker 1: and get to the point where, you know, NVIDIA's growth

Speaker 1: rate will start to level off and really mature. So

Speaker 1: I think that is the biggest risk at this point

Speaker 1: in time. But keep an eye out for AMD. You're

Speaker 1: going to see them really come on hard, I think,

Speaker 1: over the next couple of years. They're just now ramping

Speaker 1: up Helios. And we think there's some significant potential for

Speaker 1: that over the next couple of years.

Speaker 2: Angelo, great stuff. Thanks so much for joining us. Angelo Zeno,

Speaker 2: Senior Vice President, Tech Lead, CFRA Research Program.