Look, in my opinion, we're less than 15% through this AI build. And that's why, you know, I just continue to be so bullish in terms of everything I see from chips to hyperscalers to software.
Hey everyone, welcome back to Zeta Global Live 2026. We have over here one of Wall Street's massive bulls for the better part of the bull market here, alive and early. Mr. Dan Ives, 25 years on Wall Street, eight years at Wedbush, and also you got your bachelor's in Penn State. You're from...
We are huge Penn State fans.
That was a crazy way to end the season last year, but let's see.
Yeah. Look, I mean, obviously it's a rebuild year, but like we're huge fans of Campbell and everything he's building.
That's awesome. That's awesome. Yeah. So, I mean, what we were talking about outside for a little bit was basically OpenAI dropped the news today. We're recording this on Thursday, October 8th, and OpenAI basically said that they're going to be coming in $20 billion below their $70 billion guidance. Now, I know this is creating a bit of fear in Wall Street and that a little bit of downstream effects to Oracle with the delay in the New Mexico Jupiter project. What do you think about all this FUD that's actually happening here? Because respectively, and I know you probably agree with me too, is that this has been happening for a long time now. Like when are the bears just going to stop? When are they going to finally understand that we're going to see the ROI, ROIC very soon on all capital spending?
I mean, look, the cat's already out of the bag. In other words, like we're talking about what's going to be three to four trillion of, you know, of spending in terms of the AI revolution. And whether it's like circular financing, the Dario essay, this OpenAI, we're going to continue to sort of have the skeptics, and anytime you have a piece of data, they'll sort of, you know, groundswell around it. Look, in my opinion, we're less than 15% through this AI buildout. And it just, the derivative impacts, the second, third, fourth derivatives are just starting to play out. And that's why, you know, I just continue to be so bullish in terms of everything I see from chips to hyperscalers to software.
No, for sure. And early this year, not more than six months ago, just a few months ago, there was the doom and gloom that AI is going to kill software or the quote-unquote SaaS apocalypse. I know you've been a mega bull on Snowflake for a long time. How does it feel to be pounding the drum when Snowflake was less than $200 to now back even post-Slotman? Now we have Sridhar Ramaswamy at the helm. No one believed that he'd be able to do it. But now we're sitting at Snowflake near all-time highs, just pumping out 30% plus growth and expanding margins. Like how fulfilling does that feel?
Look, I mean, because I viewed it like the SaaS apocalypse is kind of more of like a fictional narrative. Now, Adobe, Intuit, like you're definitely going to have, you know, models that are definitely under pressure, but if you think about the use cases that are playing out, Palantir obviously leading it, but like Snowflake is, in my view, I actually think like ServiceNow, you know, Salesforce, many others that are going to ultimately be beneficiaries. Look, I think it's one where like Snowflake is a good example of like the narrative. Sometimes in this market they create huge opportunities. And you go back to like that SaaS apocalypse, like I always viewed Snowflake as one, like the technology, they have a huge competitive moat, but sometimes it takes investors a quarter, two quarters, three to recognize the true sort of, you know, what, the monetization of a story. And now we're getting into like the use case phase where I believe now Snowflake is going to be one of the huge beneficiaries in that.
Yeah, for sure. I agree. I mean, Snowflake, who's still the chairman, is Frank Slootman, which you probably know pretty well. Sridhar Ramaswamy came in and that was when this stock took a massive haircut after a great quarter, but a change in management never really bodes well for Wall Street.
And Frank Slootman before that, he was the one who basically took ServiceNow IPO, as you remember. And now we have Bill McDermott at the helm, who's just a dog on Wall Street. He is, he's one of the greatest salesmen I've ever seen. First of all, it's not just like Titan of industry. Like when you just go back to like the SAP days and Oracle and just like he's someone also that just understands the industry and where it's going. I think more, you know, as much as basically anyone out there, which is why he's someone like when it comes to like Trump bringing all these leaders to the White House, McDermott's there, Jensen's there. I mean, these are kind of, you know, Nadella and others, these are the leaders obviously, Karp as well, among many that are kind of leading for the first time in 30 years, US is ahead of China when it comes to tech.
For sure, for sure. And it's funny how you said, you mentioned Nvidia, Jensen Huang, who is the godfather of godfathers. I'm wondering because right now we're at the Zeta Global conference and you saw the massive excitement that came out, especially when it comes to the AIM, which is basically Athena built on the Nemotron, an open-source model from Nvidia. And we've always thought of Nvidia since '23 as the GPU play, but Jensen Huang's going out there and he's getting stake in all these small data center players. And as far as like your top pick goes when it comes to the data center NeoClouds, what is your top pick?
I mean, look, my five favorite picks for this year going into 2027, it's Nvidia, it's Microsoft, it's Apple, it's Palantir, and it's CrowdStrike. Because I just view it that it's not just about like the chips. It's not just about the hyperscalers. It's about software, cybersecurity surface area increases, and Apple, which is really going to be the toll collector on the consumer AI highway. Look, I think it's one where, like you said, the skeptics will be there, any sort of news they'll come. We're going to have like 800 to a thousand data centers get built, you know, over the next, call it, two or three years, even if you assume like 10 to 15% that get voted down. And I just think we are going into just a golden age of this AI revolution.
Yeah, and it's not even just the AI revolution, it gets translated toward the pick and shovel, which has been here for a while, and especially AMD. Like Lisa Su went to South Korea recently and she signed deals with Samsung, SK Hynix, and all these companies. It's more than just being in the US and the US onshore narrative where you basically have Amkor and TSMC building factories in Arizona. They're outreaching to a lot of the demand that we're seeing outside of the country, and you're seeing it in China as well. And I know you've visited more than a few times. Like AI is a global force. It's not just a US...
But it's a two-horse race between US and China, and then it speaks to like globally. This is just, we're in year three of a 10 to 12-year buildout, and that's also why like when it comes to things like AI safety and guardrails and different things that we see. I mean, the reality is that if we slow down because of regulatory, China wins, and they won't win in my opinion. But Nvidia's third-rate chip is two years ahead of where Huawei is. Yeah, but if regulators, the politicization, look, I think that continues to probably be the biggest risk, especially going into midterms, where, you know, relative to like data center and a lot of the grounds. But look, big tech, they've caused a lot of it. Saying like going to wipe out jobs, 50% of white-collar work, which I've always viewed as kind of more like scare tactics or whatever, but then you can get to the top, you know, top floor and then pull the ladder up.
Yeah. I mean, the way that I see it, and you probably have seen it when you were at Wedbush, and now you're probably seeing it at Yorkville Ives because you're building the tech sector for Yorkville. Is that AI is something that you need to take hold and you need to like really utilize as a tool instead of being afraid of it. And I feel like when the computer first came out, people were, they did the same thing. When the internet first came out, people were afraid, right? But look at 20 years later, we're using it at an astronomical rate at this point. In 1905, those that are like, 'Ah, this car thing, it's gonna be a fad.'
Yeah, it's about you gotta bet on the horse and buggy. 1995, typewriter, word processing companies like, 'Dude, this internet, compute, Windows thing seems like a fad.' I'm just saying like this is where the world's heading.
Yeah, for sure. For sure. So, now we have the big tech earnings coming up in just a couple weeks. We have TSMC as well as Netflix kicking it off. TSMC obviously reported their numbers as far as month-over-month goes. Probably wasn't the numbers we wanted to see, but the earnings is really what's going to matter ahead of time. And Samsung actually reported earnings recently and Wall Street wasn't too happy about it. Do you think there's some sort of cyclicality or breaking of this cycle coming up anytime soon?
We don't see equilibrium till late 2028, early 2029. The cyclicality that's maybe being like factored into memory stocks, we're just not, we're not believers in because this is a true super cycle for memory.
Yeah. No, that's true. I mean, if we really go back to 2021 when Nvidia basically topped out and you had all the copious amounts of liquidity being injected into the system, a lot of that compute that was being built out was mostly for gaming, was mostly for Bitcoin mining. But what we're seeing today is data center where shelf life could last years. And we've seen even the A100s from Nvidia still out of stock, high in demand, even increasing prices according to research from Warren Pies over at 314 Research.
Yeah, the data is there, but the bears just keep getting louder. But the bears have called 10 of the last two downturns. And they've missed every transformational growth the last 25 years. But that's what makes a market. Exactly. And like, look, are there going to be white-knuckle moments? Could there be points in the next few years that like certain stocks get ahead of themselves or the sector? Yeah. But it just goes back to like, think about just over the last year, SaaS apocalypse, software names down 50, 60%. Cybersecurity, Anthropic coming out. Look at where CrowdStrike, Palo Alto were at back in April. Like you could go through so many instances. Nvidia earnings shrugged the shoulders the last two quarters, has the stock lost its luster now? Okay. I'm just saying it's just very, very easy to be narrow. Like SpaceX, the lockup, where the stock now? Okay. No, but I'm just trying to explain like narratives are dangerous when they start to dictate and they get into like the psychology, especially when they're detached from fundamentals.
Exactly. I've always been a proponent of data and I do follow technical analysis. I'm sure you guys have done and so you guys do that in your shop at Yorkville Ives. Um, so going back to the capex segment, we were talking about earlier in the hallway how you have Wall Street consensus basically saying that Google capex is going to be somewhere in the high 200s, low 300s, which is crazy because that's almost double from what they guided for for 2026. If I look at who is going to benefit the most, I'm thinking Broadcom and Marvell Technology.
I agree. Matt Murphy went on stage the other day and he said he's looking at $20 billion for 2026 and he's looking at, I think it was somewhere around $70 to $90 billion of revenue or an opportunity by 2031. That is insane. And I do think from just the research that I'm seeing is that a lot of it's going to come from the TPUs that they're signing up with Marvell, with Google. Look, it's, and also Google is in a great spot right now relative to what they've shown on GCP and the type of growth they're seeing. Look, it's an arms race and these companies cannot slow down relative to what's happened across the sector. But I think Marvell, I also think downstream when we think about just NeoClouds, Nebius, CoreWeave, is just broader beneficiaries of what's happened on the hyperscaler side.
Yeah. And a lot of those NeoClouds and hyperscalers, not hyperscalers, but data center builders, they're guiding for in the mid-single digits in terms of billions, and these companies are not realizing a lot of revenue today, and we're coming to the point where their revenue is really ramping up very, very steeply. But going back a little bit to the return on invested capital question, actually one of my good friends, Craig, he does a lot of research on Wall Street and trades a lot of stocks as well, and he actually had a question for you. Do you mind if I ask a question?
What's your friend's name?
