I'm

long Oracle.

You can take everything I say with a grain of salt.

Someone at Oracle said, why do you always say that before you talk about us? But I want you to know I'm biased.

So I'll give you my honest opinion, but understand what my bias is.

I was worried.

I see this in Force Majeure, like, well, I hope there wasn't a hurricane in Texas.

I mean, doesn't this have tentacles that are concerning for the market?

Do you know what that means?

I do now, as do all the rest of us.

This is a great cultural moment we're having here.

We've got some California, we've got some New York Stock Exchange, we've got the Aussies.

Yes, you know I've got so much love for the Aussies, having a daughter who lived there last year.

Look, I think what's going on is a translation issue, quite seriously.

I think that Oracle's trying to tell its story to investors.

So wouldn't the sky be the limit for Oracle? Are we worried about the cash position then compared to some of the other so-called hyperscalers and the credit rating, the negative free cash flow?

Along with Anthropic, from what I read in your note, the new frontier.

Isn't that funny, right? Anthropics using Palo Alto and also creating the threats that Palo Alto has to defend against.

And Anthropics partnering with Palo Alto to try to create those defenses.

It's a very, you know, this is how arms races work.

When everyone has arms, everyone uses arms.

And you can see it both in the case of the hackers and the defenders.

You know, Palo Alto spun up this examination using mythos and other tools.

What are you thinking as far as how the market might react? Is there going to be moments here that this could be super?

What makes that moat so difficult for an AI native competitor to disrupt?

Making semiconductors is hard.

Breaking news.

Write down that I said this.

I could have said it 30 years ago, too.

What these guys do is so incredibly difficult.

And they're using AI to make it easier for the customers of Synopsys to access the 30 years of information that they have gathered in the design of semiconductors.

but but i mean do we eventually get back to the cyclicality i mean does i suppose the music has got to stop sometime corey and i just wonder how sustainable these kind of numbers are for these companies in the business of chips and i think about it in light of another company we're going to be hearing from later this afternoon and that is marvell which i'm sure you look at very closely as well because the story with this one now is how much it's changing the portfolio mix as far as the interconnectivity business and how this used to be obviously all about the chip, but now everyone's talking about how the chips talk to each other.

Is it possible that no matter what they say, short-term sort of momentums can overtake things just because of how volatile these stocks are? I mean,

we can talk about the stocks, we can talk about the business.

I mean, and your question is meant to be serious, and there seems to be a conversation out there about, well, what if data centers don't get built? What if not in my backyard happens to be a thing? What if building the people complaining about data centers or water or power usage keeps these things from being built? You just don't hear that from the companies.

I mean, every single company we talk to says they cannot build it fast enough.

They cannot get enough capacity.

They can't, their customers want more and they want it faster and they're into slowing nothing down that the financing out there, right? Every morning we're waking up to a new headline about somebody selling $20 billion worth of something, some kind of financial instrument to fund these things.

The money is rushing in.

I mean, what have you seen emerge that has been like a winning theme? Because it's frankly even very difficult to gauge what the street will like, what the street will not like in these post earnings moves that have been, I mean, punishing or very much rewarding on what seems like has been overall still really strong numbers from everything that's been reporting in AI so far.

I'm sorry.

But also in the audience were two FCC commissioners, uh, Olivia Trusty and Chairman Brendan Carr.

And what we found initially through Olivia Trusty's financial disclosure record was that she had accepted over $12,000 in tickets as a gift from Paramount to attend.

What have they said in response to this report? What has the FCC said?

FCC has said, in essence, that there's really nothing to see here, but what we learned by looking at the law and then talking to four really credible ethics attorneys, almost all of these experts, they were outraged.

[laughs] There is a requirement in the rules that if an ethics official, uh, green-lights one of these kind of forbidden type events, that they document it and document their reasoning for doing so.

Well, we asked for that documentation.

We asked many, many times for the FCC to provide it, and we're still waiting.

We're all here.

This is the place where these companies I was short and spent so much time researching.

Look, we saw companies that wanted to drive up their share price initially offer very low single-digit floats really for the first time, uh, at least the first time that I can think of, um, in, in financial markets, in IPO markets, going back even into the '80s before I was in these markets.

Um, I was just a kid, but, um, it may have happened before that, I don't know.

But, uh, uh, we did go out and do a study.

Uh, my colleague Spencer Cadigan and I, uh, created a, a s- s- he did a great job coming up with all this information on what IPOs look like today.

And look, supply and demand is a thing, right? If you o- if there is great demand and you offer very little supply, you can get a, a high price until that supply increases.

[laughs]

Tempus is still here on the fifth floor.

And full disclosure, when I was a fund manager, I was short some of those companies.

And the first time I walked into the Schwab building, I looked at the address and thought, and the Schwab Network building in Chicago.

I looked at the address and thought, why do I know this address? No way.

This is the place where these companies, I was short and spent so much time researching.

Look, we saw companies that wanted to drive up their share price initially, offer very low single-digit floats, really for the first time, at least the first time that I can think of in financial markets and IPO markets, going back even into the 80s before I was in these markets.

I was just a kid, but...

Oh, I

[laughs]

... for months leading up to the deal, and we wrote a big report about the, the, how the company works and everything.

But the market dynamics are pretty unique in this deal.

Um, we're gonna publish a study this week, uh, that's gonna show that, that the typical IPO, the last 100 IPOs or so of, of any size going back almost a year typically offer about 21% of their shares out in the float.

SpaceX offered just 4%, 4.3% for those of you keeping score at home.

So what that tells us is that they really wanted to juice the stock.

Then they also had the benefit of the indexes, uh, many of them having, uh, being compelled to buy the stock in the aftermarket if they couldn't get allocations.

Mm

Is this about Starlink cash flow? Is this about rockets? Is this about what may become of XAI that's currently hemorrhaging cash? What are the fundamentals here?

I mean, the fundamentals looking backwards aren't great.

Well, let me rephrase that.

It's amazing what Elon Musk has done.

It is amazing that they've built a private rocket company doing billions in revenues.

It's amazing that they've created a way to get Internet service into rural areas using satellites.

Their AI business has not been amazing.

Yeah.

À ce point, la valuation ici, ce que nous avons vu sur le marché aujourd'hui, évidemment, en plus de 20% depuis que le stock a commencé à marcher sur le Nasdaq, est-ce que c'est à propos de l'inflation de l'argent de Starlink ? Est-ce que c'est à propos des roquettes ? Est-ce que c'est à propos de ce qui peut devenir de l'AXAI qui est actuellement en train d'émerger de l'argent ? Quels sont les fondamentaux ici ?

Les fondamentaux en regardant en arrière ne sont pas géniaux.

C'est incroyable ce que Elon Musk a fait.

C'est incroyable qu'il ait construit une entreprise de roquettes privées qui gagne des millions de revenus.

C'est incroyable qu'ils aient créé un moyen d'obtenir des services Internet dans les zones rurales utilisant des satellites.

Le business de l'intelligence artificielle n'a pas été incroyable.

Il s'est traîné vers ce que d'autres entreprises ont fait dans ce secteur.

Cela semble être un endroit risqué pour la volatilité potentielle.

Where do we stand on something like that? You know, if

you're a customer of a company that sees that you need the product desperately and there isn't a lot of product available and you raise the price, that's acceptable here when you're trading stocks.

But if you've got a long term relationship with a business and they jack the prices up to really cost you a lot of money, which is what's happening to All of NVIDIA's customers.

It's happening to Amazon.

So what's Amazon doing? Designing their Tranium chip.

Google's getting charged a lot of money by NVIDIA.

They've trained their own TPU chip.

And where does Cerebrus fit into this story?

So how much of that actually goes to NVIDIA? What's the allocation in your outlook?

Yeah, and that's not just for six companies.

The biggest builders of hyperscalers, the Meta, Microsoft, Amazon, Google, even CorWee is in that number.

But when we look at what those companies are spending, they're spending most of the largest chunk of money.

I shouldn't even say most because that makes you start to think of things like 50%.

About 33% we think goes to NVIDIA chips or goes to CPUs.

Most of that NVIDIA right now, some of that's AMD.

But do we face risks? Is there overbuilding power constraints? Is demand going to fall short? Tell me more about that.

I mean, how important is this AMD and Meta deal in terms of reshaping the AI landscape right now?

I think it's huge, and it's huge in a lot of ways.

It's obviously huge for securing a place at Meta, one of the few companies that can really afford this kind of a giant deployment for AMD, right? AMD is now there head to head.

We know that they've been there on a product way for Nvidia, but we haven't seen it in the marketplace.

Now we're starting to see that with some moves that they've made there, moves they made with Oracle earlier.

I think that's really important.

I also think it's really important as we move from a world of training large language models to actually using them to inference.

Do you see right now certain companies having core competitive advantages and finding some resiliency in terms of this broad risk appetite that has seemingly dwindled for no reason? So I'm just curious where you find these winners and what still sticks out to you in broader AI?

Corey, what did you think of the names that have reported thus far?

Well, I tend to look at what the companies are telling us, not necessarily how the stocks are moving at that particular moment.

And there have been a lot of moments in the last year where we've wondered, is the AI trade reaching a peak? Are we in an AI bubble? Is the spending not going to continue? And how is the spending changing? And I think that what we saw from the results we've seen so far from the Mag7, let's toss out Apple.

It's a very different kind of company.

But if you look at Microsoft Meta, even Tesla to some degree, what we saw was a real commitment to spending untold billions of dollars to continue to build out AI and really no slowdown in sight.

I think one of the things we're going to see here, though, and I think it's not just, it's really our fault as people trying to understand what the businesses are doing, really, and the market trying to understand things.

Not so much the companies changing their discussion, but there's more and more discussion about the things that go around the purchase of chips, right? We've talked about the AI build as being spending money on GPUs, buying chips.

This week, when we hear from Amazon and Google, which one's going to be the one that is more surprising or makes Wall Street happy? Corey, quickly.

Can we spare up a bit of capacity to get those to China? But I also wanted to get your thoughts on CC Wei's comments about being nervous about about whether AI demand was real or not? And I don't mean to speculate because I don't like to do that, but would it have been a different outcome if, say, Jensen Huang had said that?

No, I think in both cases, you have CEOs who have been through it.

In fact, these two have been through it together for decades now, and they've got enough control of their companies, enough humility.

Humility is not the right word in both cases, but they've got enough notion that things can change and that they need to adapt their companies to it.

so um when they're placing all this these bets into one place they've been through this before they've seen it work out and they've seen things not work out and it'd be crazy not to be concerned about it but i think you can see that both companies are clearly putting their their their chips on the table and they're really uh expect that this is what's going to happen and every indication they have is that that is indeed what's going to happen that we're going to see continued fantastic growth of ai and And all hands lighten the load when it comes to whether we'll get competition, and we'll have Intel helping out, and we'll have TSMC helping out, and everybody making it.

Look, if semiconductors are going to be a good thing, if AI is going to be a good thing, we're going to need lots of companies to make those things.

I'll also say that it's really interesting to see that these increased smaller wafers, right? These new kinds of chips will use less power.

How much concern do you have any of those potential headwinds?

Corey, what do you think here? Cory Johnson.

So it's so interesting what Google's doing with AI because they've got to keep up.

They recognize that this is the future of how people gather information and use information, and yet it cannibalizes their business.

If you think about it, when you go to Google and you type in a search, what you want is the answer, not a link.

But the way they have made money over the years is by providing us links.

The links show us a place to click.

We click, Google gets paid.

What's the best AI play, in your opinion, if you put them in the one and two, three slots?

Are you worried about an AI bubble or an AI pullback for these Mag7 names? Are you actually really expecting that? When does that happen and why?

So I think the place a bubble exhibits itself is in tertiary names and in companies that have already got troubles that are trying to make up revenues that aren't there and doing lousy things like accounting fraud.

Certainly that's what we saw in the dot-com era.

I don't see that in the MAG7 names.

And in particular, I don't see that in Oracle, of course, being out of the MAG7.

But I think that what was very clear from the earnings we got from three companies, Oracle, Broadcom and Ciena.

Ciena, much smaller company, but a really important company in networking that had an incredibly bullish result at quarter this.

like it was in the tech bubble when things were 100 times plus in the P.E. ratio, right? I mean, so is tech still a thing? When the stock price goes down, the

I mean, does that necessarily speak to the reversal we saw on Thursday after what is definitely a banner report from Nvidia? I could not second and third that enough.

It's not anywhere near the most expensive stock in the group.

And it is the most profitable stock in the group.

It is a company that is generating the most free cash flow, which is the ultimate sort of sign of both stability and business success.

You know, I spend more time looking at the technologies behind these companies than the prices at which they trade.

But I have to say that, you know, when I also spend a lot of time on the financials and look at the numbers, and the numbers are just really, really solid for this company.

I want to point out, too, and I wrote this in our note about NVIDIA when the numbers came out last week, that the networking sector is something that has not gotten a lot of attention as it relates to AI in general and certainly as it relates to NVIDIA.

So as far as then speaking more on that networking aspect of NVIDIA's business, I mean, how critical was the Mellanox acquisition in positioning NVIDIA as a further dominant player in AI networking? And really, does this create a more durable moat now going forward?

You know, is it a win in the long term for all of these names?

I drove by the Rigetti headquarters yesterday, as it happens, because it's down the street from the Berkeley Bowl grocery store.

The Berkeley Bowl grocery store is four times the size of Rigetti and has greater revenues.

So, you know, this is a science project.

This is like, you know, investing in a company like this or D-Wave or something is kind of like believing in organized religion and then going to church as fried chicken, as if you're participating in the same thing.

There's nothing here yet.

And maybe there will be something from IBM.

Where would you put money to work instead when it came to tech then?

We'll start off with Salesforce as Oracle's What's really driving some of the momentum behind its AI narrative? Because I know it has not been seen as a leader in that space as it's right now concerned with its dominance and its overall sales platform that we could be eroding that real true dominance in their market share.

Yeah, it's interesting.

You know, agentic AI was just, that was all that Salesforce talked about all week long.

And it seems to be the mantra coming from CEO Mark Benioff, agentic, agentic, agentic, agentic.

I mean, it was exhausting to hear the conversations about that.

But it is truly a change in the way that their software works.

I think for those who don't use Salesforce every day, it's hard to imagine the rest of us who do use Salesforce, businesses that just simply run on all kinds of Salesforce properties.

And really, so then why are we seeing this move to the downside today?