 Welcome to this week's CUBE pod. I'm John Furrier with Dave Vellante, Extracting the Significant Noise Episode 48. Dave, great to see you. You're in your office, not in the studio. I'm in my office as well. This is CUBE pod. Not much going on this week. Not another quiet week. I mean, it's only the most important stock on the planet Earth, going supernova, going ballistic. Goldman Sachs dubbed NVIDIA, the most important stock on planet Earth, certainly echoed by the market's response, whether it's earnings yesterday. I mean, their earnings report well anticipated. You predicted it right. Congratulations, unlike other analysts who were hand-waving. Seeing as a tailwind, the paid analysts out there. But the report really pushed a huge bullish market momentum. Stocks up the day after hundreds of points. Just incredible run from NVIDIA. Not new to us on the CUBE. Obviously, we've been chronicling the software aspect of that for years. The dye was cast years ago with CUDA and all their other systems. NVIDIA is the story. Ironically, they stepped on the entire and completely put a wet blanket on Intel's big event. So really this pod, I would say, dominates through. We'll talk about some other news and AI and a bunch of other stuff going. There is a lot of stuff happening, actually. But the two seminal moments this week just really, to me, put a real stake in the ground in this moment in time that really encapsulates the hype of AI right now. I mean, it is the superhype, hyperhype, AI hype. You got the Intel big event and then NVIDIA just completely throwing shade on it with its massive numbers and the thundering lead it has. So let's get into it and so much to go in. The stronghold that they have on the segment, Dave, is pretty incredible. You had an amazing tweet storm, diva latte on X, formerly Twitter, and on LinkedIn. Also, you got some commentary. Got a great set of comments on my post. I even did a QBI post on there that summarized all the new but so many data points with NVIDIA. It's the most talked about topic in technology right now. Well, yesterday was interesting because the day before Palo Alto disappointed. You and I talked about that and it sort of sent the whole tech market down. It's certainly all those cyber stocks were off on sympathy and people I think at the time said, wow, maybe this tech bubble is bursting and that spilled over to NVIDIA yesterday. People were very cautious. I mean, stock was down. Not that the day-to-day stock movements matter, but you can't help but watch it because it reflects the sentiment in the market. I put out a tweet saying, well, hang on. First of all, I said, revenues are expected to triple, profits are expected to grow 7x. These are really inflated. I can see why people are really cautious and maybe taking some profits ahead of the print, especially after the Palo Alto news. Then I said, well, there's a two-year backlog of their GPUs. They virtually got no competition unless they have very poor visibility on what they can actually produce. You would think they're actually going to beat their expectations. They did. As you said, the stock is up over 100 points today, up 15%. The quarter was remarkable. I mean, $22.1 billion in revenue, up 265%. It was interesting that they talked about sequential growth of 22%. When was the last time you heard a company talk about sequential revenue growth? You've got to go back to the 90s when tech was booming, and they beat revenue by $1.6 billion. Then their gross margin for the quarter, John, was 77%. That's more than 10 points higher than Intel ever achieved. They're having an Intel moment. I had a road post about the Intel moment in the 90s. They're having it now. Dave, this NVIDIA story is really a blend of strategic brilliance technology, technological innovation, and market acumen. Jensen years ago set the table for this. This is not new. Again, they have economies of scale. Everyone else's point of catch up will have diseconomies of scale. As NVIDIA navigates the ever-changing landscape, it's going to be incredible to watch the impact they're going to have on the AI market. The further advancements they're making just increases their leadership. The multifaceted nature of NVIDIA's journey, it's a saga of a tech giant redefining the boundaries of what's possible. I mean, think about this. The dream is to join a startup and get paid. Now, the startup journey is you become a unicorn. You're overvalued. You do a down round. You sell the company and you still maybe get liquidity. If you were a janitor at NVIDIA six years ago, you're a millionaire right now. This is an incredible story. It's a multifaceted journey. It's a story about entrepreneurship in a big company. I think there's so much to this story on how they've just made the right moves and had the right focus. Again, this isn't a startup. This isn't like Facebook and then dominating. It's just like Facebook is a lot of NVIDIA kind of in it. You can still make money in the big companies. It's the talk of Silicon Valley right now. It's like, hey, you'll never make money to join a big company. In this case, if you joined NVIDIA six years ago, janitor, worker, engineer, manager, you're rich. Google, Facebook, NVIDIA, it's a profound accomplishment. I'm very happy for NVIDIA on that one, but also I'm scratching my head going, okay, there's still a huge gap on product. The supply chain problems, the demand for AI, is this a seminal moment? Obviously, the profound journey. Love it. Dave, what's your analysis of this as a future headroom relative to NVIDIA stock? Also, that's a stock. They're platforms. I think to assess the question, how long will this last? Is this just a bubble? Is this sustainable? Won't competition eat them alive? I think you've got to go back to at least 2006 when NVIDIA introduced its CUDA software architecture. 2006. Then the whole point of that was that the GPUs that it was producing needed specialized software, proprietary software, so that people could program for this new type of computing. This is like a long time ago. Then when all the DeepMind guys and Ilya, now you're talking still 2012 when the AI researchers started really getting into programming GPUs for building large language models. Then it took them years to perfect that. If you've ever heard Ilya talk about their journey and how size mattered. Then when they started getting video and audio, it made a difference. Then in 2019 NVIDIA buys Melanox and they tune their chips for InfiniBand. Then in 2021, in the middle of COVID, they introduced an ARM-based architecture. This is all before chatGPT was announced. The point of all this is that they have been building a mode for well over a decade and they kind of lucked out. Your kids and my kids were like, can I get a PC for Christmas with an NVIDIA chip in it? This is where their whole business was gaming. Then crypto gave them a big tailwind. Remember the crypto crash? I would heard NVIDIA. The best thing that ever happened to NVIDIA and AI was the crypto crash. All those GPUs out there got repurposed instantly for AI. That's a great point. My conclusion is we are entering a multi-decades wave where NVIDIA could make mistakes. They could screw up. They could get arrogant and shoot themselves in the foot. But if they don't, I think they're going to dominate this market. I think they'll have 80% of this market, which is going to be hundreds of billions of dollars. Let me give you my research analysis on this. I did a little quick research note preparing for this podcast. I knew we'd get into it. I got seven areas to review real quick. The financial performance overview, obviously. The strategic positioning of AI and tech with NVIDIA. The future growth projections, competitive landscape, market industry perspectives, tech buzz, the risks and considerations, and then my recommendations going forward. Then up to the conclusion. Let's just get into it. The financial performance, earnings, smashing, the results were off the charts. Year-on-year net income soared by 769%. Incredible financial performance. NVIDIA beat big time. They expected big. They got bigger. Check the box on financial strategic positioning. Don't forget the valuation. It's almost 1.9 trillion now. Incredible. Financial performance. Everyone's happy. That's also powering the stock market. NVIDIA now becomes, in my opinion, a bell weather. That's a little bit down in the analysis. Point two, strategic positioning in AI and tech. Clearly, this puts NVIDIA as the market leader in GPUs, especially with their hopper series. That was key. You mentioned an infant event. These clustered systems will define the new architecture. We've been talking about in theCUBE. The term clustered systems hasn't really been out there as nomenclature. It's been coined. It means something, but I think you're going to see that become a category. Clustered systems. Watch that space. We're going to be digging into it. We unpacked that big time at the high-performance computing show Supercomputing 23. There'll be more of that at every single event. Clustered systems. Remember that name. It reminds me of Hyperconvergence Day. We're going to see a lot about clustered systems with chips, pairing chips with software, and other things. Pay attention to that. Then just demand and the leadership. Supply chain resilience. Diverse impact on their chips. They got strong sales on all their AI chips, so well-positioned NVIDIA. Future growth prospects. He anticipated another 24 billion, this huge upside to the stock. It's unbelievable. If they can keep the resilience on the supply chain, then they're going to have the growth will kick in. We'll watch that. Competitive landscape worth bullet. Intel. Huge competition. Intel's challenged just to keep pace with NVIDIA. Intel's now in the shadows of NVIDIA, particularly in the AI and chip design. NVIDIA's advanced tech and market share keeps it well ahead of Intel. Intel's staring at NVIDIA as it pulls away. AMD. Also a key player, but still good well-positioned in the market. They're nipping at the heels. Don't count AMD out. Competitive landscape. There's demand for chips. Rising tide floats everybody, but still NVIDIA is pulling away. Market and industry perspectives. Global influence. Tech community buzzes off the charts. Totally relevant. Risks. Supply chain. Okay. And the geopolitical. Huge wild card. So if you're going to bet on the stock, we'll get to the analyst recommendations in a minute. There's two things to watch with NVIDIA. Supply chain and then the geopolitical factors. China specifically. We're going to see that play out. And then finally, my recommendation, it's a strong stock to have. I would definitely put money into it. I don't see that going away. And I would watch it and sell it. Kind of keep track of it. And finally, my conclusion is NVIDIA is now officially a bellwether in the market, both on the financial stock side, strategic positioning and their market maneuvers. I think you're going to see people copy NVIDIA around clustered systems. You're already seeing moves by Broadcom and others kind of putting chips together with software. This is to me an architectural shift. This might be one of those moments we'd look at saying, okay, NVIDIA was the pay setting standard that's going to shift this cloud next gen AI. So that's my research note on this. That's a great analysis. Just a disclaimer, we're not given investment advice. But if we were, I would say buy the dip. We're not really analysts, but we'll give you the scoop here. So I mean, of course, we're relaxed. I mean, I look at the B of A analysts. Do we have to say that when they'll give investment advice? I don't know. I think it's good cover your ass. But the B of A analysts, who I find to be one of the best in the business in semis, I was just sort of walking in to this call and I heard him on the TV saying he thinks it's got 20% upside in NVIDIA. Who knows? But he thinks it's got upside. A couple of other sort of nuggets that I pulled out of the call last night, 40%, they claim 40% of the data center revenue was for the for the year, fiscal year was for AI inference. And Jensen in the Q and a said, this is probably low. Okay, so the reason why that's important is because everybody talks about, well, look, it's not just training inference is going to dominate the market, totally agree. But then they think they, they imply that that's good news for the competition, bad news for NVIDIA. And here's why I think that's not the case. First of all, NVIDIA is doing a lot of inference. Second of all, one of the analysts on the call, CJ Muse from Cantor Fitzgerald, he asked, I thought the most important question or very important question on the call, which is, he said, with the million X improvement in GPU compute power over the past decade, did today's training clusters become tomorrow's inference clusters? And then Jensen's answer was, of course, yes, because one, those are accelerated and two, they're programmable. And then the other thing that came up on the call, which I thought was really telling is Jensen said, why do you think that every cloud player, every hyperscaler is a long gaining its depreciation of servers now to six years? It's because the general purpose CPU price performance curve doesn't warrant you having shorter life cycles. You don't need to upgrade every two years like you used to, or even every three years. But every time a new GPU comes out, the demand is there. So it's a complete shift in the buying patterns of semiconductors and data center. That's CJ comment was really interesting. I quoted it in my report, by the way I posted this morning. That was a great catch by you on that. That idea of programmability is huge. Now I want to just make an extra point on that because, you know, when we came out of, when we went into re-invent, ADF's re-invent in December, you know, when I sat down with Adam Silevski, that was the key thing we talked about was inference. You look at all the trends, training is one thing and inference becomes the killer app. We coined that term before super computing in Denver. But if you look at the cloud guys, if you think about what this AI trend will do for the cloud guys, it's incredible. I think, you know, how issue and I talked about this on a video we did this week, AI becomes a cloud dream. If you're a cloud player, AI really can help your value proposition significantly. If you look at meta, okay, what they're doing, they're going to be in the game pretty quickly. So you've got meta, AWS, Azure, okay, all positioned beautifully for AI. Google. And Google, I mean, Google's new Gemini is incredible. And they got the data with YouTube. They got all the data. Most of the data in the world is video and they've got it all. Yeah, the Gemini stuff is incredible. The Gemini pro thing is going to be pretty game changing with video. Just the context window alone is incredible. So again, all the hyperscales, the big get bigger, but Dave, what does this mean for the marketplace? Again, we've seen big incumbents in these big waves, IBM back in the 80s, 70s and 80s. They were the dominant player. But they weren't as agile. They got toppled down by the PC revolution and the networking revolution. The question is, is that I'm not sure that the big guys go away in this wave. I think they get bigger. So what is that going to mean for the competitive strategy for a startup or even try to get some white space? So if you're an entrepreneur, you're a big company, again, my point about Nvidia and a startup, there's more rich millionaires in Nvidia than there is in the startup ecosystem, generally speaking, because the market's better at Nvidia right now. And that's a big company. That's counterintuitive. If you go back to the, just a few years ago, it was cool to be in a startup. But if you're working for Nvidia right now, it's cool. Yeah. And, and, and look at who's doing all the investing in startups is the big cloud guys are doing tons of investing. But I want to, I want to land on something that you said about Adam Salipski. Because I remember when the other thing he said to you is, because we're talking about inferences, we're going to do inference going to be big, but you got to train it first. There's a lot of action around training. And I think the cloud guys are going to do really well there. And, you know, this is one of those Matt Baker moments where it's not a zero-sum game and it's not a winner take. Matt Baker, you mean from Dell Technologies? Yeah, Matt Baker from Dell. It's not a winner take all market, but I will say this, you know, the Tim Crawford, you know, in that, that LinkedIn group that we're in, she's actually really, really smart, a bunch of really smart people. Tim was rightly saying, look, there's going to be a lot of specialized, you know, use cases here and a lot, a lot of opportunity for semiconductor designers to develop those specialized chips. Totally agree. Two things. One is, NVIDIA is going to play in that game. No reason they can't. And the second is, I don't think any of those guys are going to have monopoly like margins that NVIDIA is enjoying. So they'll do great. They'll make a bunch of revenue. Dell is a great company with whatever, 30% gross margin. But NVIDIA, we're talking about 77% gross margin in their guiding that it's going to, you know, long-term, their long-term model is mid-70s. I mean, that's SAS, marginal economics. David Fleuer said to me in 2021, the post that we wrote, how NVIDIA is going to dominate the enterprise. That was May of 2021. Fleuer said to me, Dave, NVIDIA is going to be so dominant. They could give away the chips and just sell the software and still have a viable business model. And of course, they're not giving away the chips anything, but prices went up this quarter. David Fleuer is one of the best key research analysts out there right now. I think he has done incredible work on ARM, too. I mean, he used to talk about ARM for years, right? Talk about when everyone was like, go back a decade, the work we did on the chip formation. ARM, we saw that coming early. We saw NVIDIA. We saw what AWS was doing when they made that big acquisition. Again, you start to see what we're covering with. And you mentioned Gemini Pro 1.5, you start to see the formation of the infrastructure game. And if you're being the real key winners here, so if you look at the stock market pop from NVIDIA, who wins on this? Google wins. AWS wins. Azure wins. Even Oracle wins. All the big iron players, when I say big iron, big compute, big horsepower are going to win because the apps need this. I put it in perspective. You mentioned Gemini, which is Google's AI Gemini Pro 1.5. An enormous upgrade to their AI models. There's one million token context windows. Huge. The previous record was clawed 2.1 with 200,000 tokens and GPT-4 Turbo with 128,000. The difference is staggering, right? And this is just the beginning, Dave. So as the infrastructure gets better, people like Snowflake win. People who store data. So the data play here is off the charts. Again, this is really nuanced in the industry, but look at the impact of this NVIDIA announcement. It points to the entire AI systems completely being modernized. That means every single company that's got growth and or assets in this market will win. There'll also be a set of losers. So what we're going to start watching and analyzing is the winners and the losers, because there's going to be losers on this wave. People are going to be driftwood from this wave. They're not going to make it. So we're going to, this is going to be the real analysis that's going to come out of this market. So if you look at our big discussion around industry analysis and financial analysis, the winners have to be positioned because you can't fake it until you make it in this market. You don't have it. You don't have it because you got to have it to win. It's exciting, Dave. I got to tell you, I've never been so pumped. Well, you mentioned up front that the NVIDIA news kind of overshadowed the Intel Foundry Day, which I don't know if you saw that. I did, but you are all over. Give us a low down of what happened there. I mean, I wrote a post on it. It was very highly produced. I mean, it was really an incredible production. And look at, you got to tip your cap to the way Pat Gelsinger plays his hand. I mean, it's brilliant. He had an amazing cast assembled. CEO of ARM, he had the commerce secretary come in over the wire remotely, Satya participated, Sam Altman, on and on and on. And they positioned it as the world's first AI Foundry, and which is, again, trying to position TSM. And they're basically saying, look, TSM is monolithic. We've got the future with the chiplet era. And really, I mean, brilliantly packaged, but I have to say, and it gives me no joy to say this, the bottom line is still the same, John. Intel is losing share. You mentioned AMD. They're losing share to NVIDIA. ARM wafer volumes are 10x those of x86. And Pat swipes it in NVIDIA's packaging, but NVIDIA and TSM, in my opinion, are ahead in packaging for AI. Now, Tim Baharan, I'm no expert in this field, but Tim is deep in it. He feels like that's not the consensus. But I will tell you, talking to my sources and brilliant people as well, like David Floyer, that chiplet approach solves a problem that Intel can't solve. Let me ask you a question. Let me just finish. It can't make big chips. What does that mean? It's for AI because they can't get the yield. So it's a classic case, in my opinion, of they can't fix it. So they're featuring it. And I'd like to dig into that a little bit. So why don't you explain chiplets, what that means? Okay, chips. Obviously, we know what silicon chips are. And when you say big chips, put it in perspective. Let's unpack this Intel, because I do agree with you, by the way. I'm rooting for Pat Gelsinger, but just when you start to see a lot of window dressing on events, there's not a lot of meat in the bone, then you've got to kind of roll your eyes a little bit. So I felt Intel a little bit over the top, a little hyperbole there. But let's look, and I'm clear they were overshadowed by NVIDIA. That didn't help either. So let's get into this chiplets because I think that was the innovation that Intel had was their strategy to be something different. And why do you say it may not work with the versus say big chips? What are chiplets and explain? So I think chiplets are going to have their place. But by the way, chiplets are nothing. I mean, they're not new. I mean, the packaging is modern and Intel is using EUV technology, which again, they relate to adopt EUV. Everybody knows that, but they now finally getting on board and adopting advanced EUV. But it's a 40 year old technology goes back to IBM mainframes. Back then, wafers were really tiny. And the technology to make big chips that didn't exist. It was actually, interestingly, it was the consumer markets that brought big chips. And I'll explain what I mean by that was the iPhone. And as almost always is the case, it trickles in to the enterprise. But so you think about it. Think about a Grace Hopper chip. You got all these layers of memory. And the key to their architectures, if you have a huge amount of SRAM, just, you know, non volatile RAM, it sits on the chip and has all the, all the different components share that memory. So it's the CPU, the GPU, the, the language processing unit, the neural processing unit, all the accelerators have synchronous access to that memory and that big Grace Hopper chip. And that is the core of solving the AI problem, AI, big data, and big compute. And you need fast connections between all of that. That's where Infiniband comes in. That's why I don't think Infiniband's dead. We had had that little debate. It's the internal networking that Melanox brought to NVIDIA. So when you make chiplets, which is what Intel's doing, you're taking a bunch of smaller chips that with much, much slower connections between them. And they're asynchronous. And so Intel's saying they got 95% yields on their, their chips, but if they can't make big chips at high yields, they just don't have the volume. And so it's a way for them to get to market. And chiplets are going to have a place in the market. There's no question about that. But for the high end, high, when you say big chips, you don't mean, you don't mean like size, because like the iPhones have big chips on them. iPhone's a big chip or a big chip architecture. Yes, they have a big shared SRAM with NPUs and CPUs and GPUs, all sharing that resource as opposed to the, the chiplet approach, those connections are asynchronous. So it's an order of magnitude less bandwidth and slower. And it's just not the winner, in my opinion, for AI, the AI era, those so-called monolithic, which I know is a pejorative, is going to win. The trading off performance with the smaller chips, which is really relevant in AI, because those connections between the chiplets by definition are slow, those certainly slower, and they're asynchronous. So AI, AI gives monolithic the advantage. And I guess the other one more point if I may. So Tim brought up in Twitter and he's right that there's a rumor. Tim Crawford. Tim Crawford. No, sorry. Ben, I said Tim. Ben Baharan brought up and he's probably right, he's closer to it than I am, that NVIDIA is supposedly adding a chiplet architecture. They might be working with Intel to do that. And I'm saying, okay, great, they're doing it because they can. It's additive to their high end AI machines. The reverse is not true. Intel can't get to monolithic. They can't get the yields, so it can't play in the high end. Jensen's going to own that. He's going to get 80% of that market. And look it, maybe Jensen's wrong about the AI factory, but I wouldn't bet against that. I mean, I think, I comment as a re-invent when he mentioned AI factory. I think what he means by AI factory, I think he means AI operating system. These clustered systems is a real deal. I mentioned it in my report. And what you're getting at here is interesting. That's why I put up the iPhone because we used to have big chips, my mind went, oh yeah, Psy is a big motherboard like a PC. No, it's monolithic in the sense that it's controlling a lot of the parts for shared services, not trying to be a small bunch of collections. So silicon diversity is a big trend right now. But what you're saying is AI needs more horsepower with that monolithic architecture. So to me, this is an architectural debate. Is that the right approach? So my feeling right now is I believe that the current Grace Hopper clustering with Infiniband is beautiful if you're a hyperscaler. You got constraints on power. It doesn't matter how many GPUs you put in a rack, you're going to be constrained on things. But by folding them together in the clustered system is a good concept. You get more throughput. It's a supercomputer basically. But can you do that on an iPhone? Okay, so like you said, how do you make that architecture scale horizontally and also enable that same kind of power? I think that's going to be small, fast or cheaper clustered systems. So what runs all this? What software, Dave? So this is where we've been teasing out on theCUBE over the past two years and last year in particular, is there going to be a new operating system? NVIDIA could win the OS model for AI. If the AI factory has code words for, we're the next Linux for AI for metaphors. That's a great metaphor. I mean, CUDA is that OS and they've got hundreds and hundreds of libraries for very specific use cases. And again, chiplets are going to be useful. But for the most rigorous AI where shared memory is critical, chiplets are going to fall short. And I would argue that in the AI era, until I think is calling it the chiplet era, but in the AI era, bigger chips with fewer connections, I think are going to be the winner. And they're going to have the best price performance. So if that's true, let's just take our insight, just say we're connecting some dots here. If you look at their stock today, 700, pushing 800. Okay. Oh my God, the stock's too high, that's a trillion dollar valuation. No one, everyone gets scared of that. But if they truly are the operating system of AI, go back to 1988. I remember the moment when I realized that, oh my God, Microsoft's got the operating system, system software, and the application stack. That's a monopoly. Their stock price was pushing it high right then. They were already overvalued. If you had bought Microsoft in 1988 as a stock, where would you be today? You'd be pretty rich. You'd be pretty rich. We don't give stock advice. Nvidia though, could have that same vibe. So again, if you look at their track record, that's what you're looking at. If their software model, this AI factor becomes the AI operating system, then you want to bet on Nvidia. And again, the wild cards are competition, geopolitical, and supply chain. If you look at those things play out, if I was laying all my chips down, I look at those three things. And if I feel like that's going to flip in the favor of Nvidia, then their lead just extends. And then they might not be able to get caught up by anybody else, because if you're anyone else, there's too many diseconomies of scale that put you in a position of potentially faltering. You might be able to match the story with some capabilities, but to get the trajectory of that experience is just going to be very difficult. I think Amazon could be one to watch Google or Meta. Meta is with the Broadcom, Octane on the board. I think that's interesting. That's an interesting wild card. Here's an interesting thought exercise. You remember in 99, 2000, Cisco was the most valuable company in the world. And then .com blew up, and Cisco is a great company, but not the most valuable company in the world. Is Nvidia the next Cisco, or is it the next Intel, Microsoft, next monopoly operating system of the data center, as you call it? I think that might be the wrong company to compare them to. I think more like Microsoft. I think my analogy of system software and operating system is interesting because at that time, Microsoft was under the radar. They're just a PC company. They had DOS, and then Windows became the DOS enhancement. You can argue that could be Nvidia's move into AI factory. The applications are driven by the AI thirst from GPUs. They own that by default. You can call that the application layer. Then you have potentially a better analogy. All Cisco had, really, what Cisco didn't have any technology. They had a collection of piece parts that they acquired over the years. But what Cisco had that nobody had at that time was they had the position locked in everyone's routes. Routing software and routing tables were connected systems. Those connected systems were hard to replace, and switching costs were extremely high. If you had Cisco, you couldn't really switch. You'd have to shut down your company, hire a bunch of people. The costs were way too high. They had a monopoly on networking. They milked that monopoly, made so much cash, and then did a lot of acquisitions and created little moats and little pockets, but still always had that problem of being one unified thing. Cisco did a good job of keeping that going. Microsoft continued to get into the internet. They had the browser. They made some mistakes under Balmer, but then something that telecorrects everything. That's why again, the reason I bring up Cisco is because think of era. The internet era, the information highway, all that stuff. Cisco was the network and the on-ramp to the information highway, and that's why people said, okay, they were the picks and shovels, if you will. A lot of people look at NVIDIA as the pick of the shovel. I like what you're saying about, they actually lead me to say, they're the next Cisco, which is, hey, good company. Once they used to be the most valuable in the world, now they're just a good company, or maybe a better example on the other end of the spectrum is, are they the next wind tell, John? That's kind of what you're implying with Microsoft. Yeah, I mean, exactly. I mean, different markets, same generation, but Microsoft had that going into the internet. Remember that Cisco was formed in the 90s with TCP IP when the OSI model came out. What that did was it disrupted all the mini-computers who had proprietary operating systems. IBM had SNA, digital had DECnet, everyone else had other things. So that was the proprietary mode at that time, but once the standards Ethernet and then TCP IP created internet working, that category enabled massive growth, wealth creation, market growth, new companies, everything happened after that in a big way. The internet was born, everything just went supernova. So again, that was the Cisco way. Microsoft had the software side. They had the monopoly. So if GPUs is the new monopoly, then Nvidia has to create a mode, but also a sustainable competitive advantage, and the software gives it to them. They get that AI factory to become the connected system for clustered systems. Then you could argue that clusters are the new data center. Clusters will be at the edge. They'll be at the core. They'll be on-premise. So clustered systems is something to watch. And again, remember that term because I think we're going to come back to it a lot. So can you imagine, I'm always crapping on Lena Khan, can you imagine, I can't remember, we talked about this in the Q pod before, can you imagine if Nvidia was able to buy ARM? No, Lena Khan didn't kill it, but her proxy in the UK did. They were going to pay 40 billion for ARM. And ARM is today worth 130 billion and SoftBank owns like 90% of it. Imagine if Nvidia got that. So maybe they did get it right. I mean, I think that's why I think AI systems is like camouflaging their play. I mean, Nvidia right now is going to create so much awareness to their market. You know, what's the expression Jeff Bezos used to say, your margin is my opportunity. That couldn't be more true than this, but they're doing right now. You know, John, remember, Jesse said to you, he said, I'm not sure Jeff ever said that, but I don't think I think Jesse was just trying to cover for Jeff because it's been quoted, but because it's been sourced, sometimes he pretty much should own at this point because it's been super glued to his thing. But the margins of your opportunity means is barriers to entry. People are going to come in his market. And Nvidia just created a massive splash in the global financial and industry technology pool, biggest splash of all time. I mean, just today, this is Thursday of this week, the biggest rally of the year on the stock market as of right now globally because of Nvidia. This is going to be the moment we'd look at saying this is where AI really becomes, wow, super frothy, right? Are you saying, so, you know, that quote that's attributed to Jesse, you think about, you know, physical retail and how Amazon has been disrupting that. You think about so many examples, Uber, your margin is my, my opportunity to the taxi guys. Are you saying that Nvidia's 70 mid 70% gross margin is an opportunity? Well, like in my, in my seven point research note, I just read earlier in the pot, I think it's all the above. It's one, it's the financial opportunity, money to be made. That's the number one. Hey, this shit, this cash to be made here. I want that margin to leadership. So what's happening is that will this reason why there's a global stock market surge and probably a technology surge from infrastructure and chips is because everyone from Dell to HPE who makes PCs to anyone who's building any systems because remember, clustered systems will be the future. They're going to all win. So again, there's going to be a massive list of winners and then a list of losers. This is what's going to happen in the smart executives, the smart teams, whether they're a startup or a big company being the next, that will become the next Nvidia, they're going to go after that market opportunity because it's a large market. What's the old investment thesis tape go after a large market with a, with a okay product, you'll do good. Go after a small market with a great product, not so good. Here you're going to see people enter the market. The demand is going to be massive for AI systems and the operating system, the infrastructure, the apps that run on them is going to create a massive renaissance, a Cambrian explosion, wealth creation, huge trajectory of value. And this is why everyone's kind of running as fast as they can with things like RAG, retrieval, augmentation generation. That's a data play. The gen AI stack, no one has yet built, in my opinion, a full gen AI native tech stack. That's going to come. So there's a lot of things evolving so fast. It's just, it's just incredible. Back to the competition because the tech industry is there a bunch of alphas, right? And so there's, you got guys like Jassy, you got Frank Slutmans, you got Satya, you got Lisa Sue, Jay Sri Ulaal. These are all alphas. They want to be winner, right? And, and they don't like when other companies get monopolies. So they're all sharpening their knives right now against Nvidia, you know, saying that Nvidia is gouging. Don't, don't leave out China companies, either, by the way. Well, that's the other thing is China, to me, China is one of Intel's biggest competitors. Intel's fighting a multi-front war. If China, if they don't export their stuff to China, China will build their own. Oh, well, yes and no. But so maybe. Here's the thing. I forgot to mention, we were talking financials. All this, I could be not mention this. All of this Nvidia blowout goes down with virtually no China, single digit, like China contribution. Okay. And they said it's just going to be single digit going forward. So they had a big China business that is getting decimated because the US government won't let them sell. So they've come up with other products, but that's without China. Now, where China is, I think of concern for Intel and for everybody longer term, but your Intel in particulars, it's like, it's like when China got into the steel business, they're doing rebar, low-end stuff. I fear that Intel is going to, an Intel foundry is going to get stuck with a lot of the low-end processes that they talk about their advanced processes, but they'll probably suck up a lot of the low-end business. They're partnering with China, China foundry. So some of that IP is probably going to trickle. So that, to me, is a big competitor. They got AMD fighting. They got Nvidia crushing it with AI. They got ARM with all their wafer volume standards. The hyperscalers are all developing their own silicon. So Intel is just surrounded. And then they're fighting TSM and Samsung with the call becoming the number two foundry. All this is coming down to where do you run your AI? Where do you host it? Like, right now, managed services exist like open AI. The Lama and the Mistral open source models, the foundation models are growing like crazy. So you have this frothy developer environment and the underpinnings that support that infrastructure and platform software is rising up to meet the demand. So when you ask the question around long-term prospects for Nvidia and this new architecture, complex systems, distributed computing, clustered systems, the things we're kind of kicking around, you've got to look at two areas to see where Nvidia will flinch or competitors will win. And that is if there's going to be a shift in the market dynamics. If the market changes rapidly to some other position and the competition intensifies, that will be the tell sign. So if you look at the next year, is there rapid market shift? Does it stay on course? And what's the competition do? Those are the tell signs. That's what I'm looking for. If those things stay the same in this three-level stack, infrastructure, software, apps, super chips, super cloud, super apps hits, you're going to have a super cycle. I mean, that's what's happening, Dave. And I think if that stays in and Nvidia wins in the operating system, if it doesn't happen, then there's an opportunity for competitors to kind of get position in there. So this is going to be the big land grab. This is why I love your analogy of these, the Cisco and Microsoft conversation because those were the beginnings of a big market that was developing. Yeah. So let's talk about the competition. I mean, what competition is what I would say. AMD, okay. They got some AI chips coming out. Intel, I mean, you should see, watch the Intel replay. I mean, it was a big production and music and guys in hazmat suits, throwing around whatever of swag. I mean, it was a big audience every time Pat was up there. I mean, it was, I'm like, wow, okay. Big celebration. So the thing about Intel, John is by my count, they're building six fabs and Pat Kelsinger by his own admission says these things cost like 30 billion now a year. So where's Intel going to get 180 billion? They're not going to, it's not the cash flow of the business isn't going to throw that off. Are they going to take out 180 billion dollars of debt? Not likely. Are they going to get 180 billion dollars from the US government, which is 33 trillion dollars in debt? Not likely. Is Sam Altman going to write him a check for a trillion dollars? Maybe a couple hundred billion. If he gets a 7 trillion, maybe, but I really know what that's all about. So I keep saying Intel could run out of money trying to build all these fabs. Why do you think they're delaying building some of these factories? They're just out of the cash. Dave, what a great week. Big focus on Nvidia ground breaking. I mean, record setting, Wall Street, Intel, all the chips are going to win. Semiconductor is hot. Silicon is hot. The silicon angle is big. It's happening. Go to siliconangle.com. The Cube Research. Check out that site, cuberesearch.com, cube.net. Dave, we'll wrap it up. We're going to be in Barcelona next week for Mobile World Congress now called MWC with a massive set there. We've got the executives from Broadcom coming on, talking about chips. We're going to have the CEO of HPE, CEO of Dell, CEO of Cisco, see a lot of all the top executives going to be coming on. Uniper. See you next time. We'll be doing the next podcast from Barcelona. See you next time. See you guys.