My mission is simple: to make you money. I'm here to level the playing field for all investors. There's always a bull market somewhere, and I promise to help you find it. Mad Money starts now.
Hey, I'm Cramer. Welcome to Mad Money. Welcome to Cramer and welcome to Nvidia's GTC conference in San Jose, California. Other people make friends. I'm just trying to make a little bit of money. My job is not just to entertain, but to educate, to do some teaching. So, call me at 1-800-743-CNBC or tweet me at Jim Cramer.
Today, we saw the price of oil can actually go down. But will it stay down? Oh, come on. Let's not get ahead of ourselves. This market rallied today because President Trump is allowing Iranian oil to get through the Strait of Hormuz. That caused the price of crude to drop back to the mid-90s, which is all that mattered to the averages. Dow surging 388 points,
the S&P jumping 1.01%, the Nasdaq pole holding 1.22%. Now, I know we've all gotten a little Pavlov in this market. We buy stocks whenever oil goes lower, which isn't very often, and we sell stocks whenever oil goes higher, which seems to be all the time. So today people bought stocks and bought them aggressively because oil went down a fiver. How important is the price of crude to the stock market? Let's put it this way. We left here on Friday thinking the next stop for oil would be $150.
We started calculating what $150 would mean at the pump. What it would mean at the voting booth? What would it mean to stocks? Put those rate cuts on hold because of inflation. No, make that stagflation. How bad would the recession be with $150 oil? Did President Trump not know about what Iran could do with the chokehold of the Strait of Hormuz? Was this another war we're going to lose? You heard it. You heard it all.
As I surveyed the street's research this weekend, I kept seeing the same kind of notes. As oil skyrockets higher, this litany I just ran through, this full-on parade of nightmarish horribles should cause the S&P 500 to drop between 15 and 20%.
Pessimism was everywhere. Defeatism. We were back asking if the center would hold again. But then a funny thing happened on the way to $200 oil and stagflation. Oil itself didn't cooperate. Why did it go down? It's all so murky and confused. But I believe the president's decision to let Iranian oil through the strait demonstrated a level of practicality that most traders didn't think the president was good for. A level of practicality that's been sorely missing in this bombing war.
So, a new narrative suddenly emerged and caught lots of traders leaning the wrong way. If the strait is open to tankers from Iran, then how can oil get to $150? Doesn't it need the US to do something to Iran's oil to get to $200? If anything, maybe oil has overshot its destination already from the stock market's perspective.
The most dangerous day in this entire war was when Israel struck an Iranian oil facility that unleashed attacks on all the surrounding countries' oil infrastructure that cemented higher prices in our minds. But Trump has now decided to lay off Iran's oil industry and even let them export their oil. He's making a point that he isn't going to attack Iran's oil infrastructure. I'm not saying he won't change his mind. I mean, maybe he'll attack Iranian shipping tomorrow.
He quickly went back to joking about how he took out two different sets of leaders, but he seemed to be ready for whatever happens, including an end to the war. Traders seem to welcome his swagger. Why don't we do this? Let's call it a quasi off switch, called a pause button, that shocked the pessimists and just created a huge wave of stock buying.
And what did they buy? Oh, what did they buy? Well, why don't you just take a look around here? I mean, this is what they bought. They bought everything in the room. See, what you see behind me is the wide world of Nvidia. This is GTC, the AI festival of everything that drives the tech world, right here in this hall. Everything. Oh, yeah. Very all my faves.
This is what they bought. Oh, they bought hardware. They bought the software. They bought the infrastructure. They bought the data center. They bought everything that produces AI, agents of AI, leaders of AI, friends of AI, buddies of AI, accelerated computing. They bought the whole wide world of Jensen Huang and everything in it.
We went to listen to the keynote speech that Jensen Huang gives every year. He told the tale of endless demand. You know, this man previously forecast three months ago in Washington, maybe four months ago, that he had a half a trillion dollars worth of demand for his newer AI chips through the end of next year. Well, that was pretty exciting. And now he's talking, well, wait a second. He just doubled it. He said there's an even trillion dollars worth of business by the end of 2027. Yeah, he has a trillion dollar book of business.
Nvidia's stock shot up huge when that number came out like this. All right. Then it gave it up as we were getting a little sick of this, frankly. Sellers emerged. I myself was a little surprised because that new forecast of double. Nothing sees that a double. Who else has a double? I might have a double. Anybody have a double? I don't see any doubles.
I believe there will be a delayed reaction. The stock will go higher when analysts digest things because you know what? When you go from a half a trillion to a trillion dollars in sales, it does matter. Jensen told more than just a story of elevated demand. He also announced he had new chips for clients who want more inference-oriented semiconductors at a lower price.
One of the reasons why Nvidia seems stuck, just stuck at 183, if anything, is that many of its customers seem determined to make their own chips. I heard a confident Jensen talk about the chips he's making from a new acquisition, Groq, that are faster, better, and cheaper than anything else out there, including all these chips that these customers are trying to make.
There's no reason for his GPU customers to make their own inference chips if they can buy them for a lower price from Jensen's Nvidia. Again, people yawn, but they won't yawn when analysts put pen to paper. Trust me, on day one, as I strolled the floor here, I saw all the companies that are in the world of Nvidia. And I realized something. I can't name a company that isn't in Jensen's world, except perhaps this one real competitor, AMD.
Can you imagine a world where everyone's your client? Everyone, because everyone needs computing. I'd say that's pretty darn extraordinary. Maybe that's why I think this stock deserves the appellation of being the biggest.
Tonight we're going to talk to some executives who can explain why this Nvidia revolution is so important not just for the companies but for your pocketbook. Darn it. We're going to talk about how AI is going to a new level with the next generation, the Vera Rubin chip, allowing agents to reason on par with humans, in many cases better than humans.
I'm looking at a robot here. He's quoting, like, wait, the Declaration of Independence, the guy's quoting. I swear to God he's doing that right now. Now that's why we're here. It's not a sideshow. It's not an escape from this trade of Hormuz. Not that I wouldn't mind one. It's what matters when the smoke clears from Iran. It's where the earnings are. It's where the sales are. It's what moves stocks.
Nvidia is the most profitable company I've ever seen. And it's still at 183. Enough already. I know a lot of you aren't willing to think past the war. I get that. But if you're going to manage your own money, I'm urging you to look further into the future. Look behind me for heaven's sake. Or let me do it for you.
When this market can look past Iran, it's going to buy AI and everything in it because this is the future. You just can't see it because so many things are going wrong in the world. But it's only when things go wrong that the bargains surface. Here's the bottom line. Stop, listen, and learn. The opportunities are all around me. I want to share them with you over the next two days at the Woodstock of AI.
That was a really big festival from when I was younger. It's the world of Jensen Huang. It's GTC. I say we take calls. I say we go to Amos in California where I am. Amos,
Hey Jim. I have coffee with you every morning. And I'm concerned with the tariffs and over Pfizer. It pays a great dividend. And I was just wondering what your take was on it.
Okay, Amos, Pfizer is not an espresso. Pfizer is more like Nescafe, but that's okay. It's at $26. It yields 6.4%. It's got a lot of stuff in the pipeline. I am with Amos and tomorrow morning when we have coffee, we're going whole milk. Trust me.
Now, if you're going to successfully invest in a time like this, you got to look past the warden of places like where I am right now. I'd like to circle around but it would probably strangle me. At least let me look around for you if you will.
Tonight, Synopsys, one of the latest recipients of the Nvidia investment. So how is the company pushing the boundaries of AI innovation? Let's find out with the CEO. Then another one that has just been stuck endlessly and is driving me to make my head pop. You know what I mean? I mean pop like a squeezed pimple. I'm talking about ARM Holdings.
It's quickly becoming one of Nvidia's most important partners. I'm learning more about what made ARM the partner of choice when I speak to the company CEO. And how has the legacy tech company Dell been reinventing itself and powering its stock higher? I'm getting all the latest with the CEO. So I say stay with Cramer.
Don't miss a second of Mad Money. Follow Jim Cramer on X. Have a question? Tweet Cramer #madmentions. Send Jim an email to
[email protected] or give us a call at 1-800-743-CNBC. Miss something? Head to madmoney.cnbc.com.
Since we're at Nvidia's GTC event, it's time to take a closer look at some of the AI kingpin's most important partners. Take Synopsys. That's a leader in electronic design automation for semiconductors and electric components. It's now a big player in engineering, simulation, and analysis following its acquisition of Ansys last year.
Not only are they a key part of the chip food chain, but they also rely heavily on AI and accelerated computing to help their clients design better products. Now, the stock's come in pretty significantly since last summer. This is off 35% from its high last July. So, we got to ask ourselves, is this a buying opportunity? Earlier today, we got to speak with Sassine Ghazi. He is the president and CEO of a very exciting company, Synopsys. Take a look.
I feel like we have to reintroduce you. In the time since we saw you last, Synopsys has become a bigger, better company.
Yes. Actually, when we met last time, we were just announcing the acquisition of Ansys. Since then, we closed the acquisition even though it took us about 20 months to do so. It expanded the opportunity in a significant way. Synopsys' roots have been in silicon design. Now it expanded to physics. Physics is important for physical AI and how to model the environment, the world and product. So exciting opportunity.
I think that self-driving cars may be the biggest use of AI. I think that answers itself, maybe the centrality of trying to figure out how to make it work. I mean, for instance, let's say you want to try to figure out black ice, how cars work on it. You can crash a million cars or you can use you and save a lot of money.
Exactly. So, think about the future products. They're going to be intelligent systems where AI is infused everywhere. They're powered by silicon and tons of software. How do you do this in a physical prototype or a physical testing? It's impossible. It takes too much time, too expensive, etc. So, what we do, we virtualize the chip. We virtualize the actual mechanical system with Ansys and we model the world with the partnership with Nvidia Omniverse. They have the ability to model the world. We bring in the physics fidelity and accuracy to create those products.
One of the reasons why your company stock has been so amazing is you are the linchpin between Nvidia, say, and Taiwan Semi. You literally can't just have Jensen Huang call Taiwan Semi and say, look, here's what I'm cooking up, but they can call you because you are a very important part of the phase before it gets to Taiwan Semiconductor.
What's often not understood is the leading semiconductor companies, they cannot just architect the chip and assume it's going to be manufacturable. We're the bridge in between. There is something called design-technology co-optimization. You have to go between here and here multiple times and that's what we do as a company. We provide the design solution between an architect and manufacturing at the chip level.
All right. So let's talk about prototyping and digital twin. Why do we need to have a digital twin? Why can't we just put up factory after factory? What do you save with digital twin?
The cost of iteration. Because if the product is not complex, of course, you don't need to go through the effort. But as you're envisioning those future complex products, especially with AI infused in them, where they have to operate in a real-world, uncontrolled environment, they have to reason, learn, act. How do you do it with a physical prototype? You have to virtualize every aspect of that stack from the silicon all the way up to the software that's running on it.
Well, this is beyond the 10th grade physics that a lot of people took, obviously. Now, if I were to go to your booth, please explain what happens. I mean, in the end, this is a bit of a trade show. I don't mean to reduce it by saying that, but you've got product that people can buy.
Yes, of course. At the silicon level, if you're a semiconductor company, we have the products that our semiconductor R&D engineers use to envision the product, to implementing the product to make it manufacturable. With the Ansys acquisition, it opened up our market significantly. If you're designing a jet engine, an arm or a hand for a robot, autonomous driving, etc., you need to simulate, you need to make sure whatever you're envisioning is going to be manufacturable and it's low cost doing it at a schedule that is practical. So that's what we're doing with our customers.
You hit upon a term, low cost. I mean, the fact is that it may take thousands of laborers to do what you do and it's too expensive to explore, to create in this new world unless they have you. And there's a shortage of engineers to do it right.
Today, you cannot get enough engineers to design and deliver to these products. Actually, today at our booth, we have companies like Applied Materials. We have companies like ADI, Analog Devices. We have a company like Honda. We have a number of companies that are participating with us to say we cannot enable that future without having a virtualization of these end systems and we need the technology, the solution from silicon to system to deliver to it.
Well, this is so important, Sassine, because right now what we hear about, oh, they're going to be giant layoffs. How about the fact that we couldn't do the things you're talking about because there aren't enough people? You substitute it without you. Is any of this stuff cheap enough to do?
We are racing ahead with bringing all kinds of AI and agents into the workflow because our customers don't have enough resources or time to deliver to these chips. Think about chips. Three, four years ago, it was 18 months to 24 months to deliver a chip. Right now, it's down to 12. Imagine the inference if it's operating in the physical world. That design cycle has to operate at 12 months or less. How do you do it and deal with that complexity without that automation you bring in?
And I don't want to be too pedestrian, but the fact is the stock did fall in part because of just delay of approvals. To me, it wasn't because of any shortfall. It was just because you had to get the deal done and now people are getting the opportunity to buy the new expanded Synopsys.
Yeah, we had a restriction to sell to China for six weeks. That was a significant headwind. True. We were right in the moment of wrapping up the acquisition and as you know, we don't run the company and build the portfolio on 90-day increments. We have to look at it for the long term and we cannot be more excited about the opportunity.
Well, I can't stress enough that investors should be thinking in the long term because that's how you've built this company and it is an amazing company and with Ansys, it's even better.
Thank you. I mean, when you look at Synopsys, a 40-year-old company, Ansys, a 50-year-old company, we have an amazing opportunity.
It's just terrific. Okay. Sassine Ghazi is president and CEO of Synopsys, SNPS, long a company we have championed since the beginning of Mad Money. Sassine, thank you so much.
Thank you so much. Thank you.
Coming up, even the likes of Nvidia needs production support sometimes, and the CEO of one of their biggest helping hands is sitting down with Cramer next.
As we celebrate all things AI, at Nvidia's annual conference, I want to highlight a company that's become increasingly important to the entire semiconductor industry: ARM Holdings. This semiconductor design company cut its teeth providing CPU designs that maximize energy efficiency, especially for smartphones. That expertise has made ARM the preferred partner for many AI systems, including Nvidia's.
That said, for some reason, the stock has been stuck in a rut, trading sideways for the past couple of years. Can it get its mojo back? Because I got to tell you, it's got the earnings back. Let's take a close look with Rene Haas. He's the CEO of ARM Holdings to find out. Mr. Haas, welcome back to Mad Money.
All right, so Rene, I want to start by saying I look around, you look around. Does anything not have ARM Holdings in it?
I would say probably not. There's a whole bunch of robots that just fired on here before our show. Those robots are ARM inside, I'd say virtually every single one of them. Obviously all the smartphones are, but the reason is that ARM is the mobile technology that now powers everything around AI. So I doubt you'd find anything that did not have ARM inside on this show for it.
Okay. So Jensen is GPU. That's a different kind of chip. Many people felt that he thought at one point that we would not really need a lot of CPUs. That's no longer true. We know he needs a ton of them. And you're the one he has chosen to make them.
There was a time when this AI thing sort of kicked off that everything was all about the GPU. And that was really because the tokens, the information was being generated by GPUs. But you still needed CPUs to work in orchestration with that GPU. That was Grace, ARM-based. That was the first generation Nvidia built. And then Vera, which is your next generation system.
What's happened though, Jim, over the last year or so is that as everything moves towards agentic AI, in other words, agents running the work, agents run much faster than humans. They do jobs. Those jobs need to be done by CPUs. So what we're seeing is a huge explosion in CPU demand. And even Jensen today announced the Vera standalone system. So we announced the standalone system on CPUs which of course are ARM-based which indicates the demand for CPUs is now moving to a complete different level.
Now I think that people should have to know when we think about robots, we should be thinking not just Optimus. When we think about the fact that people are saying self-driving cars, well, self-driving cars need agents because they would use too much power otherwise. A lot of this has to do with efficiency, the idea of using your chips. So saving energy is just incredibly important and it's you.
There's a lot of ARM inside these data centers, but there's a lot of ARM, maybe even more, in these mobile systems such as autonomous vehicles or robotics. And why is that? All of that work around the AI workloads has to happen locally inside the machine. You can't have latency in terms of a car braking or a robot making a maneuver that has to rely on the cloud.
So AI now that finds itself into physical devices all has ARM inside because of two reasons. One, we're the most power-efficient system in the world. We're 50% more power-efficient than the competition. And of course, table stakes. All of these machines have to have CPUs, which is why ARM is the CPU of choice there.
Do you think that there's still this hangover of people who think before you buy ARM, how are cell phones doing? Because I see them as an increasingly smaller part of your mosaic.
You know, in the last earnings call, I said that in a few years, and my IR guys were really careful that I use the word few versus a numeric number, data centers would be our largest business. No question about it. That is going to happen very, very quickly, largely driven by all this demand for power efficiency. Smartphones, great business, not going away anytime soon, but the growth for us is really around physical AI and data center driven by power efficiency. And yeah, you're right, people may have overlooked that, but we've reported the numbers consistently that that's where we're heading.
Right. Can you help me with where you are in terms of scale out? What does it mean when a lot of Jensen's GPUs are together and they got to talk to each other but they can't, they don't know how, it seems like they push it to you.
Yeah. There's a lot of buzzwords out there, right? Scale up, scale out. What does all that mean? When you have a single rack and there's lots of GPUs and CPUs in that rack and they all connect to each other, that's mostly us. That's scale up. All those racks now to be connected to each other, that's scale out. All of that work in terms of managing the storage, managing the data, managing the interconnect, that's all ARM. So, we're seeing an explosion of demand. And thank you for asking that question of ARM usage in the scale up and scale out domains largely because you need power efficiency to go manage all that traffic. There's a lot of traffic.
And how about in a car? What really does happen if you have a self-driving car? And we're going to be able to make it so that every car is going to be self-driving. What does that mean for ARM's business?
It's going to be gigantic already. We are the default inside the car. You look at your digital dashboard. That entire display, that's all ARM running the display that now integrates with the autonomous agents inside the car that ARM drives. Again, similar to a data center where tokens are generated. The CPU is the brain, the control panel that runs everything inside the car and in the robot. So physical AI 10 years from now, maybe less, might even be bigger than data centers for us.
You know, I'm kind of mystified that your stock isn't going up more as this total addressable market is so much bigger than when I first met you. I mean, I remember discussing what happens versus Qualcomm. I mean, you've kind of left them in the dust. You also have, there are other companies that do CPUs but they are not close to Jensen like you are.
You know, when we met two years ago, the company was going public, people did not believe that we could grow 20% year-on-year for two or three years and we said we're going to be able to do that. We've exceeded that. I can tell you honestly, Jim, and I was, you remember when we met, I was so optimistic about the growth of the company. Two years later, I'm far more optimistic today than I was two years ago.
The line of sight we have in terms of our future business is so strong in terms that we can see two years out, three years out, five years out in terms of demand and everything is moving our way because the CPUs are needed for everything and in a bigger way. And what's the biggest thing they need? Power efficiency. And we're the best at that.
Now people always ask you about what SoftBank is going to do with its stake. I mean, when I listen to you, I say I don't really care. Well, there's something in the industry and I've learned as being a public company CEO about hedge funds, short-onlys, long-onlys. The ultimate long-only is Masa. He has not sold a share.
And there was a lot of rumors and buzz that when the company, when SoftBank was doing these investments in OpenAI, etc., Masa was going to need to sell his shares. And I said to you on our last earnings call, he has no intention of selling. And he said, and you can tell the folks I said that he is the true long-only.
Now, one thing that I want people to understand, and you helped us with scale out, I need you to explain something else to people because it's why I'm so bullish on your company. There are 22 million developers who have chosen ARM. That means once they choose you, it's kind of for life. It's really sticky. This is really the ultimate endorsement of ARM.
The magic of a hardware business is the software that runs on it. The more developers you have, the more software that's written. The more software that's written, the more hardware there is. The more hardware there is, the software guys say, why would I try to design software to anything else? So, it's this flywheel that flies and gets bigger and bigger and bigger and it's a very, very hard momentum thing to move the other way. It's been going our way.
Now, people should know you have a long relationship from Jensen dating back. You guys work together. I mean, to me, it's very clear that another company that does make CPUs is an opponent. So it really is. And now he likes Intel too, but there's AMD and AMD I don't see as competitive right now in a lot of these markets that we're talking about.
