Nvidia’s “Burn the Ships” Moment

In this podcast, Motley Fool analyst Tim Beyers and host Ricky Mulvey discuss:

• Highlights from Nvidia‘s earnings report.
• What future supercomputing systems could create.
• CrowdStrike‘s first quarterly call with analysts since the outage.

To catch full episodes of all The Motley Fool’s free podcasts, check out our podcast center. To get started investing, check out our beginner’s guide to investing in stocks. A full transcript follows the video.

This video was recorded on August 29, 2024.

Ricky Mulvey: The most important company on the planet reported. Let’s get into it. You’re listening to Motley Fool Money. I’m Ricky Mulvey. Joined today by Tim Beyers. Tim, I feel like you’re on one today. How are you doing? What’s going on?

Tim Beyers: Most important company on the planet. You sure you want to go there?

Ricky Mulvey: Six percent of the S&P 500 and it’s basically been driving all of the returns. Not all of it but a significant majority of it. A significant amount.

Tim Beyers: Well, you better name that company then.

Ricky Mulvey: NVIDIA. They just reported.

Tim Beyers: They did.

Ricky Mulvey: If you look at financial media, NVIDIA is the most important company on the planet right now.

Tim Beyers: I would not use the financial media as my barometer of what is the most important company in the world but I will grant you that that is a fact. The dumbest, most hilarious thing I have seen so far in an earning season is there was an NVIDIA watch party. There was a watch party. Can we just take a moment to recognize that there was a watch party in New York City for a thing you can’t watch, for an earnings call? Now, that, what you end up watching, apparently is like the CNBC on-screen reaction. Have we lost our collective minds?

Ricky Mulvey: Kevin T. Dugan, staff writer at the Intelligencer went to the watch party. I think they only existed in New York City. When the earnings missed, he said, “A man in a colorless Cilidon button-down balked,” when I asked him his thoughts about it, looking as if I killed his dog. Others were far less attached. ”I really don’t know anything. I barely care,” one said. ”I just wanted to see what was happening with this meme event.” Tim, what’s your reaction? Barely care, or do you look like I just killed your dog given the earnings report?

Tim Beyers: Neither. I am not attached to the NVIDIA results. I think they’re interesting. I think they were good. I don’t think there was an earnings miss here. I think that the guidance wasn’t what everybody would have hoped for for a massive beat. But let’s just be clear. Revenue up 122% up 16% sequentially. Data center revenue up 154%. This is still an outstanding business, and even for the forecast, Ricky. For the Q3 forecast, there is a significant deceleration in revenue growth down to 79% but that’s still 79%. Year over year that they are forecasting for Q3. Come on. These are good results. But that is not the issue. I got sucked in so I didn’t lead with my strongest statement, so now I owe one to Matt Greer here. My strongest statement is this here, Ricky. NVIDIA just gave you its burn the ships moment. This is a burn-the-ships moment for NVIDIA, where Jensen Huang has said unequivocally, these tailwinds in AI, this growth that we have been showing you is going to continue for a long time. The way I know that is because NVIDIA just agreed to do a not 10 billion, not 20 billion, not 30 billion, or even 40 billion, but a $50 billion stock buyback. What the hell is happening? They do not have $50 billion, Ricky.

Ricky Mulvey: They don’t, because this is a $3 trillion company. This seems like chump change, Tim.

Tim Beyers: No. (laughs) Fifty billion dollars is never chump change. In terms of net cash, NVIDIA has about, what is it, a little more than 26 billion in excess cash, which is a lot. But now you’re saying, we need roughly double that in order to fund these buybacks. Now, during the quarter, NVIDIA will tell you, and they will be right that they generated meaningful cash flow. I will say, even if you strip out everything, all the debt repayments, all the stock-based comp, all the share repurchases, any cash acquisitions, and the CapEx, you still get $3.4 billion in free cash flow. If you’re to annualize that, you’re looking at $13.5 billion in free cash flow. Clearly, they can fund it if they keep growing at this rate. If Jensen Huang is right, my point is, Ricky, is that he is telling you, he’s not making any room for there to be a cresting wave. Hey, we’re going to ride this for as long as we can, because it ain’t going to last forever. He is not saying that. Not only is he not saying that, he’s going all gas no brake, and saying, this is going to keep going for way longer than you think it is. If you think this is going to slow down and hit a bump in the road, you are wrong. That is my interpretation of what Jensen Huang is saying here.

Ricky Mulvey: Also, maybe Jensen Huang thinks the stock is undervalued. If they’re putting that many billions of dollars into stock buybacks. I’m getting a shrug of Mogi from Tim for those listens.

Tim Beyers: That does not seem likely but keep going.

Ricky Mulvey: Let’s get into some more of the actual business highlights. This is really a data center business and data center revenue making up about 90% of total revenue, 26.3 billion on the quarter. Tim, that is up more than 150% year over year. A lot of that is coming from inference from AI systems doing inferencing. You can think of this as drawing conclusions from datasets. The simplest version of that is an email system, guessing if a message coming to your inbox is spam. Another highlight that analysts were watching is that it’s New GPU. Blackwell is in a sampling phase, which summer taking is a shipping delay. Anything here really stand out to you?

Tim Beyers: I don’t know that it’s a shipping delay. But again, here’s what Colette Kress who was the CFO said during the call. She said, Blackwell production ramp production meaning we’re shipping those chips out. Blackwell is the newest most advanced of the AI chips that NVIDIA produces. In some ways, this is what you would call a system on a chip. A system on a chip is lots of components, all the things that might typically be separated on a motherboard, but they are baked into the chip. It’s a system on a chipset. Blackwell production ramp is scheduled to begin in the fourth quarter, not the third quarter that’s coming, fourth quarter, and continue into fiscal year 2026. In Q4 we expect to get, and this is the key here.

This is why I say it’s burn the ships. We are setting expectations very high. We expect to get several billion dollars in Blackwell revenue, really, because you haven’t sold anything yet, so you’re going to get several billion Ricky. We’re not going to scale. We’re going right to it. All gas, no break. I think that is interesting, possibly reckless, but let’s give them credit because they keep delivering. I’m not going to call it reckless but again, this is regardless of what you think of the valuation, what NVIDIA is saying is that everything’s great. The demand is really high, and it is not going away. Here’s another example here. They’re sampling but what they’re saying is, don’t worry about it. As soon as we hit Q4, we’re going to have several billion dollars of new revenue coming in.

Ricky Mulvey: I think there’s, I want to bring to scope, the raw power of what they’re building. Jensen Huang talking about the Blackwell system, basically making aggregate systems out of these chips that can process 259 terabytes per second. The rough calculation I did Tim got me to about 1,400 hours of streaming 4K video per second. Or if you’re doing uncompressed blueray quality DVD, that’s 350 hours of uncompressed video per second. That’s a lot of processing power that’s about to become available. You like a reckless prediction. Any reckless predictions for what can come out of that much processing power?

Tim Beyers: Well, anything that is an extremely hard processing problem. If you were to do things like advanced simulations that are required for things like aerodynamic design. I could see things for even space travel, you’re talking about the planetary simulations and astronomical simulations for those kinds of calculations, that serious business, so that compute would be highly useful in those situations. Another that would be a bit more common that I think would be really useful I don’t mean this in a political sense, but I know the current president, President Biden wanted to do what’s called a cancer moonshot. If you’re going to do something like a cancer moonshot, where you want to deconstruct and try to solve cancer forever, that presumably is a very difficult mathematical, computational problem where we’re going to address therapies to solve cancer forever. You would think something like that. Like in the biotech industry, where you’re doing just crazy simulations. You’re working with a lot of data. You could see that but I would guess it’s specialist use cases. The dark side of it is and this is one of the things that worries people about quantum computing that when you’re able to process an extraordinary amount of data at a much faster pace and much more accurately, that is how you crack secure systems. Not great.

Ricky Mulvey: With quantum computing, I’ve heard I think this was in Michio Kaku’s book where to send secure data, you’d have to do it with laser beams because you’d have these quantum computers able to crack basically any code because they can just throw infinite data at it.

Tim Beyers: That’s the dark side of it. If you’re able to process that much information, when does it even matter? Can you actually do digital encryption and get away with it or is anything that’s digitally encrypted is essentially an open book because you have enough compute to unencrypt it?

Ricky Mulvey: Let’s talk about the stock for a second. Stock is down about 4% on the earnings results. No. Still up on the month. The headlines I’m seeing, Tim are NVIDIA tumbles after disappointing forecaster. NVIDIA pays price of lofty expectations. Another shrug emoji. Right now NVIDIA is trading at about 40-ish times forward earnings. That’s in line with AMD and not historically out of line for the company. How lofty are the expectations for NVIDIA at this point from its investors?

Tim Beyers: There’s two things that are true here and I know I’ve said this many times. Financially, you could make an argument for the stock and you just did it there on a Fard earnings multiple. I could also say 32 times revenue is completely outrageous. I could say that as well. But I think the other argument that I will make is that what NVIDIA is setting up for is an expectation of not just outrageous growth over the next couple of years. What NVIDIA is setting up for and the way that Jensen Huang is talking and their actions around 50 billion in stock buybacks is that this is a company that’s going to grow at an accelerated pace for well beyond the next 10 years. You have to decide, do you believe that? Do you believe that NVIDIA is a company for the next 20-30 years to be the company that provides the core engine of compute for the future? Because that’s the expectations that they themselves are setting. Financially, you can make an argument over the next couple of years. Sure. Take a small position. But are they going to fulfill what Jensen Huang is telling you? I think that is arguable. Now, I’ll give you one last thought on this. Why would you set expectations that high? I just want to give you a quote from Jensen Huang and tell you my interpretation of what this means, because it does explain. It answers your question here, Ricky. This is from Jensen Huang. The first person to the next plateau gets to introduce some revolutionary level of AI. The second person who gets there is incrementally better or about the same.

The ability to systematically and consistently race to the next plateau and be the first one there is how you establish leadership. What Jensen Huang has told you there, in my opinion, this is my interpretation, Ricky. Two things. First is, we got to be first. We have to be the pioneers. We want to set the pace. This is our mark to lose. The second is we know people are coming for us. There are people coming for us, and so if we don’t do that, we’re cooked. He’s not ever going to say that, but that’s my interpretation of it.

Ricky Mulvey: We’re going to get to CrowdStrike earnings in just a second. But first, we’re going to do a quick add. Ricky Mulvey with Motley Fool Money here to tell you about the Range Rover Sport, the vehicle that leads by example. Visceral, dramatic, and uncompromising. The Range Rover Sport offers powerful on-road performance and commanding all-terrain capability. Once you hit the accelerator, you’ll feel this vehicle’s acceleration. The Range Rover Sport combines dynamic sporting personality with the peerless refinement that you expect from Range Rover. This vehicle has a pretty incredible optional feature, massage seating. You can enjoy a dynamic drive in total comfort in a 22-way adjustable and ventilated front seat with massage function. It heats up in the colder months too. You’re not just going where you need to go. You’re working out some knots along the way. Look, the third-generation Range Rover Sport is the most desirable, advanced, and dynamically capable yet. I drove one. It was a great experience, and that’s why I’d like to invite you to visit landroverusa.com to learn more about the Range Rover Sport.

Tim, this is the first earnings call for CrowdStrike since the outage. Interesting moment for CEO George Kurtz. Starting with an apology, a thanks to the cybersecurity community. Laid out how CrowdStrike is changing its content release process. Before we dig into the earnings and before we dig into how that process is changing, what was your high-level reaction to CEO George Kurtz’s response?

Tim Beyers: I thought it was good. I thought there was some humility there. I thought there was some recognition of the damage that CrowdStrike had done. There were some moments where I thought he didn’t take a victory lap, but he did equivocate a little bit. He said, for example, in some cases, we responded within minutes or things like that. Don’t say that. Nobody needs to know that. We know that there were some moments that you did that. What everybody remembers is the days of blue screens of death all over the world, so just don’t do that. But other than that, I thought he was more humble. I thought that was good. From a financial perspective, you could argue that this isn’t so surprising that the quarter was pretty good because the outage came at the end of the quarter. But even the guidance for the quarters ahead honestly wasn’t that bad. Their earnings per share guidance for the next quarter is way below what the street had forecast. I can’t say that’s too. That’s not really unexpected, Ricky. They did say that over the next two quarters, they expect roughly about $60 million in lower subscription revenue as they reformulate contracts, and they make good on promises that they had to customers. They’re just trying to make some customers happy. Overall, I would say they did pretty good this quarter. But there’s a caveat there that we can get to, and I’ll leave you with this piece, and then we can follow up on it however you want to. It’s good but the way that CrowdStrike does business, where the enterprise sales cycle is so long, it could take a while. It could be well beyond two quarters from now that we really start to see the impact of this outage.

Ricky Mulvey: I think there’s got to be some discussions happening between IT departments and C-suite executives. How about moving to a different vendor after this outage? That’s going to take a while. One of CrowdStrike’s main competitors is Palo Alto. Do you think Palo Alto is picking up a lot of business because of this? Do you think for existing CrowdStrike customers, this is talk, but no action? What do you think is happening?

Tim Beyers: I don’t think there’s going to be a lot of pickup for Palo Alto, because Palo Alto, in and of itself has a platform that is stitched together, and so I think they always have a bit of a hard sale, particularly against CrowdStrike. I’m not sure that Palo Alto is the big benefactor here. I will say that SentinelOne, which is a direct competitor, that is trying to do a similar platform sale, that has for years, long before AI was the crazy it is now, they talked about an AI-infused platform for endpoint protection. We use it here at the Fool. It’s a good product. I had been bearish on SentinelOne for years, Ricky, because just the cost that they incurred to acquire a customer was so outrageous. It just looked like they were never going to catch CrowdStrike. For me, it’s going to be really interesting to see over the next four quarters head to head. How do I compare what SentinelOne is doing versus how CrowdStrike is doing? If SentinelOne is materially lowering its cost to acquire per customer, while CrowdStrike is having a harder time keeping customers, that will be interesting. I don’t think it’s Palo Alto. I do think SentinelOne is going to be one to watch. But I’m still not sold on SentinelOne. I should be clear about that. Really, they have lit so much cash on fire, Ricky that I really need to see them sober up before I’m genuinely going to get interested in SentinelOne.

Ricky Mulvey: Let’s pretend for the sake of this discussion, you’re a CrowdStrike customer and you’re hearing from George Kurtz, this is what’s new. You’re going to be able to basically choose when content is deployed when you get an update. We’re doing these quality assurance enhancements, including getting independent vendors to review the code before we send it out. Also, we’re not going to roll out all of our new content all at once, so you don’t have that widespread outage. Are you satisfied with that?

Tim Beyers: Part of me would be furious and part of the reason for that is you didn’t give me fine-grain control. I think in this case, let’s be clear. CrowdStrike isn’t saying, hey, you get to defer. Not getting an update. If you don’t want an update, we’re not going to give it to you. They’re not going to do that. They’re not going to be negligent. But what they will say is you get to choose. Let’s use an example of where things happen. Let’s take Delta, for an example, or Delta had all of those blue screens of death at its customer agent terminals, which really screwed them completely. You could have Delta say, hey, look, we will not turn on a distribution of the CrowdStrike update to those agent terminals until we, here at HQ, push it to some other environment, we run a few more tests through it. It doesn’t push out to the entire network automatically, Delta says, there’s an update. We put it to this testing environment first, then we’ll test it, and then if it works for us, then we’ll let it go out to the agent terminals.

It’s basically they don’t get to choose when the update comes. What they do is they get fine-grain control. You’d have to imagine that there are some chief information security officers who are like, I’m furious that you didn’t let me do this before because this was basic. That is good but you can imagine that just knowing that these were basic things would create a point of frustration. But overall, is it good that they’re doing it? Yes. It is absolutely good that they’re doing it. By the way, they are not the only ones. Tim and I are going to talk about this on this week in tech tomorrow. They’re not the only ones that have to be involved in this conversation. So does Microsoft. Interestingly enough, I am glad that these two recognize that. There will be a summit in Seattle on September 10th, where Microsoft is bringing in all of its partners that have that very low-level access like CrowdStrike does to come in and say, let’s talk about this, so we don’t take out 8.5 million Windows computers again. It’ll be CrowdStrike, Microsoft, and presumably, several others to talk this through and come up with better processes. That’s good for the industry.

Ricky Mulvey: Last time you talked about CrowdStrike, I think it was with Mary. It might have been Dylan. Usually, those are not the hosts that people get confused on the show. Anyway, you were in a wait-and-see spot with the stock. How about now? Any questions remaining? What’s your vibe check on it?

Tim Beyers: I’m still in a wait-and-see mode, but I will say that one of the things I was looking for is good numbers around net new ARR, so that is new annual recurring revenue, essentially, new contracted revenue coming through the door. It was up to 218 million in the last quarter, that was up from the prior quarter of 212 million. I need a full quarter of the post-outage CrowdStrike to see what that net new ARR looks like and I’ll be very curious. If they can keep roughly around the same level or not lose too much, I’ll be very encouraged. I thought this was a decent quarter. It looks like they’re moving in the right direction. It looks like they’re taking steps to go in the right direction. I thought it was awesome even for those who missed it, I don’t know what we would call. I think it was like the Pond Awards, where either the CFO or the chief product officer. Essentially, it’s like the tech industry, the security industry’s version of the Razi. The winners of the most horrible movie gets the Razi. This was it, the biggest security failure and it went to CrowdStrike. CrowdStrike sent a senior executive to accept the award for the most boneheaded, biggest security mistake. I thought that was a good sign, too. It does seem like they’re moving in the right direction, but I’m a wait-and-see guy here.

Ricky Mulvey: Tim Beyers, I appreciate you being here. Thank you for your time here insight.

Tim Beyers: Thanks, Ricky.

Ricky Mulvey: As always, people on the program may have interests in the stocks they talk about, and the Motley Fool may have formal recommendations for or against. Don’t buy or sell anything based solely on what you hear. I’m Ricky Mulvey. Thanks for listening. We’ll be back tomorrow.