Elon Musk, Chip Giant?

0:01Travis Hoium:Do we have a new chip company in town? Welcome to Motley Fool Hidden Gems Investing.

0:09Travis Hoium:Welcome to Motley Fool Hidden Gems Investing. I'm Travis Hoium joined today by Dan Caplinger and Tim Beyers. And guys, I think we've got to start with what I think was one of the wildest news items of the week. TSMC has been this break on this AI CapEx spending for years at this point. But now Elon Musk is saying, hey, we need more chips. We need more capacity. I am willing to put up to$119 billion into becoming now not just an EV company, not just a satellite company, not just an AI company, but also a chip company. Tim, I want to start with you. What in the world is going on here? Because it seems like chips is one of these businesses that is really hard.

0:56Travis Hoium:I mean, Intel has not gotten this right over the past decade and suddenly Musk comes in and do we have a new player in town in the chip game? It's a floor wax and it's a dessert topping, you know, for those who remember their 1970s SNL references. I mean, it's, it's no, I mean, look, I've got to look that one up. There is, there is nothing that Elon Musk thinks he can't do. And so here we are now to be fair to him, what he believes is that in order to serve all of his various businesses, he needs compute. And he needs compute at scale. He needs it for SpaceX. He needs it for Tesla. He needs it for XAI.

1:34And so in order to solve that problem, he wants to create TerraFab. And TerraFab is this idea of just this massive scale chip manufacturing facility organization. I think there are reasons to ask questions, but there is a logic to it. So let's give them credit there. But TSMC is an incredibly efficient supplier. And they are able to manufacture chips at scale because they've been doing it for decades. And they do it at the smallest possible form factors. And if you want to do what Musk wants to do, which is manufacture AI compute, then what you end up getting is a real push to manufacture the smallest chips possible right now that's at two nanometer.

2:22DSMC is already there. Musk is presumably going to get there. But I think there is a lot to be determined with this. The vision from a logic, as with most things that Elon Musk gives us, there is a logic behind this, but the difficulty level is just extreme.

2:42Travis Hoium:Dan, the reason that this is, I think, is so important is it impacts so many companies in the entire ecosystem. You've got the chip manufacturers, TSMC, you have Intel, you have the AI chip suppliers like NVIDIA, AMD. Is this something that investors should be thinking about? That maybe there's a new supplier in town, that maybe this is going to be this vertically integrated company that's going to take on a Google? I mean, where does your head go? Because it seems wild on the surface. But like Tim said, there is some logic to it. So you got to take it seriously. There is. But I think the thing you have to remember is when it comes to Elon Musk, Musk is going to look after Musk's own companies first.

3:25I think that the whole purpose of this, regardless of talking about overall industry capacity, regardless of saying, hey, there might be an opportunity for us to make a contribution to the industry writ large. Really, I think the TerraFab vision is about helping SpaceX, helping XAI, helping Tesla to do what they need to do, to build Optimus robots, to make massive launches of goods and services into orbit to further Mars missions, solar power capacity for space-based data centers, even the chips necessary to get all the way to full self-driving, which we're still waiting for on Tesla's. That is the first and foremost there.

4:10And because the trend was always for Musk to try to vertically integrate that, I don't think that Taiwan Semi or any of the other chip fabs really need to be terribly worried about this because from a long-term standpoint, it was going to happen anyway.

4:28Travis Hoium:So you think that Musk building out this chip capacity just makes sense in the long-term Musk vision? And I want to put some numbers to this. SpaceX is not public yet, but it could go public later this year and definitely is a stock that a lot of people in the Motley Fool universe are probably going to at least be interested in. But Tesla is going to be at least a partial funder of this tariff fab if it ends up happening. They have$44 billion worth of cash right now, but they also have almost$8 billion worth of debt and their free cash flow. I believe they're going to spend$25 billion this year in CapEx is likely to go negative this year.

5:06Travis Hoium:You also need a lot of money to be a car company. So, you know, Dan, the thing that we have to throw into this too, is that this is the vertically integrating is very risky because if you don't have the demand, then you just are stuck with a whole bunch of fixed costs and a whole bunch of CapEx that you've spent. By the way, we found out this week that XAI, which bought all these NVIDIA chips, didn't have demand for their Grok AI tool. So now they're just selling that to Anthropic. That may end up working out. But if we get to a world five years from now where you're suddenly building all this CapEx and we do have some sort of bubble, that seems like a really tough place to be.

5:46It is. But I think that Tesla in particular and Musk more generally has generally been pretty good about adapting to changing market conditions. The example I'd throw out is Supercharger. I think at some point Musk would have thought that Supercharger was going to need to compete against other third-party charging infrastructure. But the fact is that the third-party charging infrastructure just has never come close to matching what the Supercharger network was. And so given that opportunity, given excess capacity in the Supercharger network, given excess capacity and demand from the rest of the EV world, Tesla saw an opportunity.

6:27Musk saw an opportunity opening up essentially the supercharger network to these third parties. And I could see the same thing happening here. I think the question you raise is, Travis, is Musk going to get the timing right or is he going to make these supply agreements too late? There's a lot in the air right now. And it's going to take three or four years even for this project even to come close to completion, let alone getting up to capacity. So you're right. The timeline is going to be important. And we just don't have a lot of our information about this is coming from a one-page property tax abatement application in some county in Texas.

7:05So we don't have the details yet. But we've been around the block with Musk plenty of times. The details come when they come, and you kind of fill in the blanks. In the meantime, there's a whole bunch of people trying to speculate about how those blanks are going to get filled.

7:20Travis Hoium:All right, Tim, let's speculate a little bit. When you think about the investment landscape and this kind of an announcement, the one company that came to mind for me is, I mentioned early on, TSMC has kind of been this break on the AI build out. You can't overbuild when you can't get enough chips. Right. We've now seen Intel, their stock has skyrocketed, which maybe gives them the cash and the ability to invest a little bit more. We've seen, you know, Micron and other memory companies say that they're going to build more capacity. And now you have Musk coming in as well with$119 billion plan.

7:56Travis Hoium:And so that's an area that I'm curious to hear what you think about the supply-demand dynamics longer term. But is there any other areas for investors for either opportunities or risks as a supplier like this potentially comes into the market? OK, well, let's let's talk about how this probably plays out and then I'll give you a thought. I think the first thing is Musk is, you know, he's done some crazy things. He's not stupid. I think he might be what my friend Bill Mann likes to sometimes call Harvard stupid, where he's so arrogant that he thinks he can't be wrong. I think that is a feature of Elon Musk.

8:39But he knows enough to know that he can take this in stages. And I do think there is a stage where TeraFab can come in and solve a problem. And that's with memory. I don't know that they're going to go straight to the most advanced chip making, Travis. I don't think they would be smart to do so. But if you are filling part of the capacity constraints for memory, that would be super interesting. And who do you partner with to get there? I'm not entirely sure who that would be. Wouldn't surprise me if it was like a Samsung. But you mentioned like who is, where do you go for investing in this type of opportunity?

9:19I'll give you two that I think are interesting. One is you've already mentioned, which is Intel. And I know the stock is up materially, but the stock is up materially on Intel foundry doing nothing. Bupkus. It really hasn't done anything.

9:34Travis Hoium:They're the reported partner with the SeraFab, with the 14A, which isn't actually operational yet, but that's the idea, I think. Right. And so you're talking about, so 14A, now it's not exactly this, but that is now you're getting under that sub two nanometer. And so it does make sense. Like if Musk wants to compete at the most aggressive form factors to make the most aggressive compute, and if Intel can actually pull this off and have a manufacturing process down way under two nanometer, then yeah, that is an interesting partner. And it could really scale up Intel Foundry. And if Intel Foundry becomes meaningful, billions and billions of dollars of revenue and profit, then suddenly the math on that company looks a little different.

10:22The other, I would say, is ASML because you cannot get away with building all of this stuff and not having the monopolist for the most advanced chip making equipment. If you want to build...

10:38Travis Hoium:Isn't that a break on these plans, though? So, I mean, are they sold out for the next however many years? A hundred percent it is. And there really is no sense that there is going to be an alternative to ASML. And I don't think Musk is going to build it. But ASML right now is in a wonderful position for the next five, maybe even 10 years. So, look, I don't think you're going to get unbelievable returns on ASML. But if you or I could get 7 % over 10 years and that was relatively guaranteed in a market as turbulent as this, would you take it? I would. So I think ASML is interesting. Well, it definitely is going to be interesting to watch because we have the SpaceX potential IPO coming in the next couple of months still, I think, in the plans for June.

11:33Travis Hoium:and what happens with Tesla? Does that get rolled into SpaceX? Do they build out a new chip manufacturing business? Lots to watch and we'll be covering it here. When we come back, we're going to talk about the death of the SaaSpocalypse. SaaS stocks actually went up this week. We'll tell you what's going on. You're listening to Motley Fool Hidden Gems Investor. In a world full of noise, long-term thinking stands out. On the Capital Ideas podcast, Capital Group leaders explore the decisions that matter most in investing, leadership, and life. It's a rare look inside a firm that's been helping people pursue their financial goals for more than 90 years.

12:11Travis Hoium:Listen to the Capital Ideas podcast from Capital Group, published by Capital Client Group, Inc.

12:19Travis Hoium:One of the down parts of the market in 2026, there's been some hot spots in semiconductors, in energy, but software, SaaS stocks. We've talked about the SaaSpocalypse on this show. And the potential disruption or the idea of disruption has hit a lot of those stocks. But this week, that narrative was really broken, Tim. What earnings reports stuck out to you? Because it seems like almost every company that reported earnings this week did not report something that filled that SaaSpocalypse narrative. Well, not all of them. But I think it was probably better than expected. And I think Datadog is the one that stands out for me.

12:57So that's ticker DDOG rallied over 30 % on results for Q1 that were just outstanding, much better than forecasters expected. Revenue was up over 32 % and passed a billion dollars in a quarter for the first time. It was really impressive. This is also a very efficient company. Like for every new dollar of sales and marketing that they invest, they got back$3.74 in revenue. That's the kind of thing that boosts margins and gets you free cash flow, and they are still producing free cash flow. So I think that's good. But the real story with Datadog here, Travis, is that they have some AI-powered products.

13:39They have one that observes the behaviors of GPUs. For those who don't understand Datadog, what Datadog does is it looks at the infrastructure that you have as a company. It looks at how the software is behaving, how it interacts with systems, and it looks for outliers. And when it spots outliers, sometimes it takes actions, sometimes it reports on it. A lot of the output of it is like logs, like have a log that says, hey, all this crap happened. And that is actually really, really useful. So useful, in fact, that they said, this is from CEO Olivier Pommel, one of the co-founders, and apparently he went on CNBC and said, we got seven and eight figure deals with two of the world's largest technology companies and it's for their AI research labs.

14:30Now that sounds like, you know, you are hitting all of the buzzword bingo when you say that on CNBC.

14:38Travis Hoium:So of course, and that's actually valuable buzzwords that is, you know, trying, that's like, that's not coming out of MBA school. Like I No. That's what people want to hear. So it's a healthy business that is growing fast, has good margins, has free cash flow, and is doing deals that seem to the point that you made about, hey, maybe we're breaking the SaaSpocalypse thesis here. it feels like, oh, wait a minute. This is a company that has AI tailwinds. And I'm just going to maintain here, Travis, that I still think enterprise software is the single greatest distribution mechanism for AI technology and features.

15:20I think that is true. It doesn't mean there isn't going to be disruption. There will be disruption. But if you are an enterprise software company and you have AI that is delivering for customers and you can charge for it, I think you're probably better off than the market is giving you credit for. Datadog proved that this quarter. Dan, what stuck out to you? I took a look at Digital Ocean Holdings. That's tickered DOCN. That stock was up between 40 % and 50 % this week. They announced their latest quarterly results on Tuesday. And they had a bunch of AI related news that helped build sentiment.

15:57They launched their AI native cloud product during the quarter. As a result of that, they saw recurring revenue from their segment of AI specific customers triple year over year. They posted 22 % sales growth in the first quarter of 2026, but they now expect that 2027 revenue growth will accelerate to greater than 50 % because of its embrace in AI. And I think to Tim's point, one of the most interesting points inside of the report was DigitalOcean said that they saw by far their biggest growth come from customers that are spending at least a million dollars on the platform. They saw a 78 % increase in the number of million-dollar customers that they have.

16:47That, I think, is important because the SaaSpocalypse kind of assumes that those big clients are the ones that have the most financial incentive to kind of do things on their own.

16:59Travis Hoium:Yeah, they're going to vime code something to basically replace their suppliers. Exactly. And what DigitalOcean is saying is that, if anything, the opposite is true. It's the biggest customers that are the most likely to value the company-specific AI integration that a digital ocean, that a software-as-a-service company is doing. that I think is the takeaway for SaaS companies is build those relationships with your strongest customers because they're the ones that are going to be able to let AI evolve with you, kind of partner with you to kind of take the best from AI rather than becoming what shadow competitors.

17:45Travis Hoium:Yeah, Tim, we've got a minute left. It seems like the story that I'm gathering from earnings season, Zeta Global is another company that I follow. Revenue is up 50%. Another AI tailwinds. The story here is investors have to think about, is AI naturally something that is a tailwind for them? Or is AI something they're trying to shove into a product that isn't really natural? And I'm thinking of ServiceNow and Salesforce are companies that are talking a lot about AI. They haven't quite gotten the same reaction. Yeah, that's true. And you have to wonder, it's the difference between an organic, innovative, like it comes from the ground up versus a bolt-on.

18:24If it's a bolt-on, it doesn't really work. Now, what both ServiceNow and Salesforce have working for them is data inside their systems. And you can build around that. But it is going to be interesting, Travis, to see not only the startups that are building from the ground up, but also the incumbent companies who are building new features from the ground up. I think the latter is going to be really, really interesting to watch.

18:48Travis Hoium:When we come back, we're going to get into a time capsule and see what Dan and Tim think about the next 10 years. You're listening to Motley Fool Hidden Gems Investing.

19:01Travis Hoium:There are moments in life that reveal who we are and who we're meant to become. For those born to lead, such moments call for a vehicle of equal distinction. Dynamic by design and uncompromising in execution, the Range Rover Sport was engineered for those rare individuals who demand the world and possess the conviction to claim it. The Range Rover Sport commands attention wherever it goes as every detail has been engineered for impact. This is the most advanced Range Rover Sport yet, filled with innovations to keep you connected, including an elegant 13.1-inch touchscreen that lets you seamlessly navigate and control vehicle systems.

19:38Travis Hoium:You'll enjoy interior refinements like sculpted 22-way heated seating with a massage function, ensuring comfort for every journey. With nearly unlimited ways to personalize, from unique colors and finishes to wheel options, you can make it truly yours. It offers a powerful drive with peerless refinement, combining ultimate luxury and unbridled agility. Exclusive offers available now. Explore further at RangeRover.com.

20:11Travis Hoium:we like to have a little bit of fun in this segment and i want to get dan and tim to jump in a time capsule and see what the future is going to look like a decade from now also look at this is what is going to be in the next big thing graveyard this is the challenge with investing with technology if you get the big things right you can make a lot of money but if you get them wrong, you can end up owning stocks that go to zero and you're wondering why that's the case. So Tim, I'm going to have you go first. 10 years from now, are we going to be still be using LLMs? Is a product like ChatGPT or a Claude going to look similar like it does today?

20:51Travis Hoium:Or is that going to be a thing of the past? No, I think it's going to be totally different. By the way, what I love about this, there's something internally we do on the investing team. it's something that Bill came up with. I'm giving Bill way too much credit right now. So this is very uncomfortable for me. This is very, very uncomfortable for me, but I will give him credit on this. We call it the DeLorean exercise. So you called it a time capsule, but it's the DeLorean exercise from back to the future where you go to the future and then you work back to the present. And so if I do that into the DeLorean here, and in this case, I believe that what we see from these LLMs right now is nothing like what we'll see 10 years from now.

21:31Usually, technology over time abstracts, and the abstract layers get more and more abstracted. In this case, I think they get more and more embedded. So you're not conversing like you do right now. It is entirely embedded inside the system doing something. it is intelligent, it is acting with AI, but it is embedded to an experience. So let's say you want to rent a car, certainly there is AI going to be involved in that, but your interaction will be, say, the voice interface, or it'll be the phone, it'll be a chat, but you're not really... So is Siri in a really good spot then? That's what comes to mind.

22:12Maybe. I could see that, but I just think embedded, not explicit, I think is the big takeaway for me.