Jensen Huang – TPU competition, why we should sell chips to China, & Nvidia’s supply chain moat

We've seen the valuations of a bunch of software companies crash because people are expecting AI to come out of type software. And there's a potentially naive way of thinking about things which is like, look, NVIDIA sends a GDS2 file to TSMC, TSMC, those logic dies, it builds the switches, then it packages them with the HBM that SK Hyde NX and Micron and Samsung make, then it sends it to an ODM in Taiwan where they assemble the racks. And so NVIDIA is fundamentally making software that other people are manufacturing and if software gets commoditized, this NVIDIA gets commoditized. Well, in the end, something has to transform electrons to tokens. That transformation, there's no, the transformation of electrons to tokens and making those tokens more valuable over time. I think that that's hard to completely commoditize. The transformation from electrons to tokens is such an incredible journey. And making that token, it's like making one molecule more valuable than another molecule, making one token more valuable than another. The amount of artistry, engineering, science, invention that goes into making that token valuable, obviously we're watching it happening in real time. And so the transformation, the manufacturing, all of the science that goes in there is far from deeply understood and is far from the journey is far from far from over. And so I doubt that will happen. We're going to make it more efficient, of course. I mean, the whole thing about NVIDIA, in fact, the way that you framed the question is my mental model of our company. The input is electron, the output is tokens. That is in the middle NVIDIA. And our job is to do as much as necessary as little as possible to enable that transformation to be done at incredible capabilities. And what I mean by as low as possible, whatever I don't need to do, I partner with somebody and I make it part of my ecosystem to do. And if you look at NVIDIA today, we probably have the largest ecosystem of partners, both in supply chain upstream, supply chain downstream, all of the computer companies and all the application developers and all the model makers and all the AI is a five layer cake if you will. And we have ecosystems across the entire five layers. And so we try to do as little as possible. But the part that we have to do as it turns out is insanely hard. And I don't think that that gets commoditized. In fact, I also don't think that the enterprise software companies, the tools makers, most of the software companies today are tools makers. Some of them are not. But some of them are workflow codification systems. But for a lot of companies, they're tool makers. For example, Excel is a tool. PowerPoints a tool. Cadence makes tools. Synopsis makes tools. I actually see the opposite of what people see. I think the number of agents are going to grow exponentially. The number of tool users are going to grow exponentially. And it's very likely that the number of instances of all these tools are going to skyrocket. It is very likely the number of instances of synopsis design compilers going to skyrocket. And the number of agents that are going to be using the floor planners and all of our layout tools and our design, design role checkers, the number of agents that are today were limited by the number of engineers. Tomorrow those engineers are going to be supported by a bunch of agents. They're going to be exploring out the design space like you've never seen before. And we want to use the tools that we used today. So I think tool users are going to cause these software companies to skyrocket. The reason why it hasn't happened yet is because the agents aren't good enough at using their tools yet. So either these companies are going to build the agents themselves or agents are going to get good enough to be able to use those tools. And I think it's going to be a combination of both. I think in your latest filings it was, you had almost $100 billion in purchase commitments with people foundries, memory, packaging, and then seminalysis has reported that you will have $250 billion of these kinds of purchase commitments. And so one interpretation is, Admidias mode is really, that you've locked up many years of these scarce components that are, you know, somebody else might have an accelerator, but can they actually get the memory to build it? And this is really Admidias big mode for the next few years. Well, it's one of the things that we can do that is hard for someone else to do. The reason why we could, we've made enormous commitments upstream. Some of it is explicit these commitments that you mentioned. Some of it is implicit. For example, a lot of the investments that are upstream are made by our supply chain. Because I said to the CEOs, let me tell you how big this industry is going to be. And let me explain to you why. And let me reason through it with you. And let me show you what I see. And so as a result of that, that process of, of informing, inspiring, aligning with CEOs of all different industries upstream, they're willing to make the investments. Now, why are they willing to make the investments for me and not someone else? And the reason for that is because they know that I have the capacity to buy it, buy their supply, and sell it through my downstream. The fact that Admidias downstream supply chain and our downstream demand is so large, they're willing to make the investment upstream. And so if you look at GTC and people are marveled by the scale of GTC and the people that go, it's a 360 degree, it's the entire universe of AI all in one place. And they're all in one place because they need to see each other. I bring them together so that the downstream could see the upstream, the upstream could see the downstream. And all of them could see all the advances in AI. And very importantly, they can all meet the AI natives and all the AI startups that are all being built and all the amazing things that are happening so that they could see firsthand all the things that I tell them. And so I spend a lot of my time informing directly or indirectly our supply chain and our partners in our ecosystem about the opportunity that's in front of us. Most of my keynotes, some people always say, "Gents, in most keynotes, is one announcement after another announcement, after another announcement, after another announcement." Our keynotes are, there's always a part of it that's a little torturous in the sense that it almost comes across like an education. And in fact, that's exactly on my mind. I need to make sure that the entire supply chain upstream and downstream, the ecosystem, understands what is coming at us, why is coming, when is coming, how big is it going to be, and be able to reason about it systematically. Just like I reason about it. And so I think the mode that you describe it, we're able to, of course, build for a future. If our next several years is a trillion dollars in scale, we have the supply chain to do it. Without our reach, the velocity of our business, just as there's cash flow, there's supply chain flow, there turns. Nobody's going to build a supply chain for an architecture. If the architecture, the business turns as low. And so our ability to sustain the scale is only because our downstream demand is so great, and they see it, and they all hear about it, they see it all coming. And so that's, it allows us to do the things that we're able to do at the scale, we're able to do. I do understand more concretely whether the upstream can keep up. For many years now, you guys have been two X-ing revenue year over year, you guys have been more than tripling the amount of flops you're providing to the world year over year. And two X-ing at the scale now, it's really incredible. Exactly. So then you look at logic, say, you're the biggest customer on TSMCs, N3 node, and you're one of the biggest scientists, N2, AI, and as a whole this year is going to be 60% of N3, it's going to be 86% next year, according to some analysis. How do you two X, if you're the majority, and how do you do that year over year? So are we in a regime now where the growth rate in the AI compute has to slow because of upstream, do you see a way to get around these, you know, how do we build two X more fabs year over year, ultimately? Yeah, at some level, the instantaneous demand is greater than the supply upstream and downstream in the world. And it could be at any instant, at any instance, we could be limited by the number of plumbers, which actually happens. The plumbers are invited to next year's GTC, you know, by the way, great idea. But that's a good condition. You want, you want to market, you want an industry where the instantaneous demand is greater than the demand. total supply of the industry. The opposite is obviously less good. If we're too far apart, if one particular item, one particular component is too far away, obviously the industry swarms it. So for example, notice people aren't talking very much about co-offs anymore. And the reason for that is because for two years we swarmed a living daylights out of it, and we doubled, doubled, double on several doubles, and now I think we're in a fairly good shake. And TSMC now knows that co-offs supply has to keep up with the rest of the logic demand and the memory demand. And so they're scaling co-offs and they're scaling future packaging technologies at the same level as a scale logic, which is terrific because for a long time co-offs was rather specialty. And HBMM were rather specialty, but they're not specialties anymore. People now realize they're mainstream computing technology. And then of course, we're now much more able to influence a larger scope of our supply chain. In the past, in the beginning of the AI revolution, all the things that I say now, I was saying five years ago. And some people believed in it and invested in it. For example, Sanjay and the micron team, I still remember the meeting really well. Or I was clear about exactly what's going to happen and why it's going to happen. And the predictions of today, and they really doubled down on it and we partner with them. Across LPDDR, across HBM memories, they really invest in it. And it obviously has been tremendous for the company. Some people came a little bit later, but now they're all here. So I think each one of these generations, each one of these bottlenecks gets a great deal of attention. And now we're prefetching the bottlenecks years in advance. So for example, the investments that we've done with Lumentham and Coherent and all of the silicon photonics ecosystem, the last several years, we really reshape the ecosystem and the supply chain, silicon photonics. We built up an entire supply chain around TSMC. We partnered with them on Coop, invented a whole bunch of technology. We licensed those patents to the supply chain, keep it nice and open. And so we're preparing the supply chain through invention of new technologies, new workflows, a new testing equipment, double-sided prototype probing, investing in companies, helping them scale up their capacity. And so you could see that we're trying to shape the ecosystem so that it's ready to supply chain, so that it's ready to support the scale. It seems like some bottlenecks are easier than others. And so scaling up co-os versus scaling up. I went to the hardest one, by the way. Which is? Plummers. Yeah. Yeah. I actually went to the hardest one. Yeah. Plummers and electricians. And the reason for that is because, and this is one of the concerns that I have about the dooms, describing the end of work and killing of jobs. And one of the things that that if we discourage people from being software engineers, we're going to run our software engineers. And the same prediction 10 years ago, some of the dooms were saying that we're telling people, whatever you do, don't be a radiologist. And you might hear some of those, some of those videos are still on the web. Radiologies is going to be the first career to go. Nobody's, the world's not going to need any more radiologists. Guess what? We're short of radiologists. Oh, but okay. So going back to this point about, well, some things you scale, other things like, how do you actually get, how do you actually manufacture two XD amount of logic a year? Ultimately, that's bottleneck. But memory and logic are bottleneck. But you, how do you get to two XD as many UV machines a year? Yeah. You're over year. None of that. It's impossible to scale quickly. You just need to, you could do all of that is easy to do within two or three years. You just need a demand signal. It's not, once you, once you can build one, you could build 10, and once you can build 10, you can build a million. And so these things are not, not hard to replicate. How far down the supply chain do you go? Do you go to ASMR and say, hey, if I look out three years or now, for me, for Nvidia to be generating two trillion a year in revenue, we need way more AUV machines. And some of them, I have to directly, some of them indirectly, and some of them, if I can convince TSMC as ASMR will be convinced. And so that's, that, you know, we have to think about the critical critical pinch points. And, but if TSMC is convinced, you'll have plenty of AUV machines in a few years. And so none of that, my point is that none of the bottlenecks last longer than a couple of two, three years. None of them. And meanwhile, meanwhile, we're improving computing efficiency by 10X, 20X, in the case of Hopper to Blackwell, some 30, 50X. We're coming up with new algorithms because CUDA is so flexible, we're developing all kinds of new techniques so that we drive efficiency in addition to increasing capacity. And so, so there, those are those are things that none of that worry me. It's the stuff that's downstream from us. Energy policies that prevent energy from, you know, you can't grow, you can't create, you can't create an industry without energy. You can't create a whole new manufacturing industry without energy. We want to re-industrialize the United States, want to bring back chip manufacturing and computer manufacturing and packaging. And we want to build new things like EVs and robots and we want to build AI factories. And you can't build any of these things without energy. And those things take a long time. But more chip capacity, that's a two, three year problem, more co-os capacity, two, three year problem. Interesting. I feel like I have guests tell me the exact opposite thing. Sometimes, I don't, in this case, I just don't have the technical knowledge to adjudicate, but well, the beautiful thing is you're talking to the expert. Yeah. Okay, I want to ask about your competitors. Yeah. So, if you look at TPU, arguably two out of the top three models in the world, Claude and Gemini, we're trained on TPU. What does that mean for Nvidia going forward? Well, we have a very different, we build a very different thing. You know, what Nvidia built is accelerated computing, not a tensor processing unit. And accelerated computing is used for all kinds of things. You know, molecular dynamics and quantum chromo dynamics. And it's used for data processing, data frames, structured data, unstructured data. It's used for fluid dynamics, particle physics. You know, in addition, we use it for AI. And so accelerated computing is much more diverse and although AI is the conversation today is obviously very important and impactful, computing is much broader than that. And what Nvidia has done is reinvent it, reinvent it the way computing is done from general purpose computing to accelerated computing. Our market reach is far greater than any TPU, any ASIC can possibly have. And so if you look at our position, we're the only company that accelerates applications of all kinds. We have a gigantic ecosystem. And so all kinds of frameworks and algorithms all run on Nvidia. And because our computers are designed to be operated by other people, anyone who's an operator could buy our systems. Most of these home built systems, you have to be your own operator because we're never designed to be flexible enough for other people to operate. And so as a result of the fact that anybody can operate our systems, we're in every cloud, including Google, and Amazon, and, you know, Azure, and OCI, and right. And so whether you want to operate it to rent or operate it, if you want to operate to rent, you better have large ecosystem of customers and many industries that be the off-takers. If you're operating it, if you want to operate it for yourself, we obviously have the ability to help you operate yourself, like for example, for Elon with XAI. And because we could enable operators in any company, in any industry, you could use it to build a supercomputer for scientific research and drug discovery at Lillie. And so we can help them operate their own supercomputer and use it for the entire diversity of drug discovery and biological sciences that we accelerate. And so there are just a whole bunch of applications that we can address that you can do so with TPUs. Because Nvidia's built CUDA as a fantastic tensor processing unit as well, but it does every lifecycle of data processing and computing and AI and so on and so forth. And so our market opportunity is just a lot larger. Our reach is a lot greater. And because we have such a large, we basically support every application in the world now, you could build Nvidia systems anywhere and know that there will be customers for it. And so it's a very different thing. This is going to be sort of a long question, but you know, you have spectacular revenue. And this revenue is mostly, you're not making 60 billion a quarter from Pharma and quantum. You're making it because AI is unprecedented technology that is going unprecedentedly fast. And so then the question is, what is best for AI specifically? And I'm not in the details, but I talked to my AI researcher friends and they say, look, when I use a TPU, it's this big systolic array that's perfect for doing major smart supplies, whereas a GPU is very flexible. It's great when you have lots of branching, when you have irregular memory access, but these, you know, what is AI? Just like these very predictable matrix multiplies again and again and again. And you don't have to give up any die area for more of schedulers for, you know, switches between threads and memory banks. And so the TPU is really optimized for the majority, the bulk of this growth and revenue and use case for a compute that is coming online right now. Yeah, I wonder how you reacted that. Matrix multiplies is an important part of AI, but it's not the only part of AI. And if you want to come up with a new attention mechanism, or if you want to disaggregate in a different way, if you want to come up with a whole new type of architecture altogether, for example, you know, hybrid SSM, if you want to use a, you want to create a model that that that fuses diffusion and auto regress of somehow, you want an architecture that's just generally programmable. And and we run everything you can imagine. And so that's the advantage. It allows for invention of new algorithms a lot more, a lot, a lot more easily. And so because it's a programmable system. And and the ability to invent new algorithms is really what makes AI advance so quickly. You know, TPUs like anything else is impacted by Moore's law. And we know that Moore's law is increasing about 25% per year. And so the only way to really get 10X leaps, 100X leaps is to fundamentally change the algorithm and how it's computed every single year. And that's MVDS fundamental advantage. The only reason why we were able to make black well the hopper 50 times, you know, I said it was 35 times. And and when I first announced it was going to black well is going to be 35 times more energy efficient than hopper. Nobody believed it. And and then and then Dylan wrote an article. He said, in fact, in fact, I sandbagged it's actually 50 times. And you can't reasonably do that with just Moore's law. And so the way that we solve that problem is new out new models, M O E's, paralyzed and disabrogated and distributed across a computing system. And without the ability to really get down and come up with new kernels with CUDA, it's really hard to do. And so the combination of the programmability of our of our architecture, the fact that MVDS in extreme co-design company where we could even offload some of the computation into the fabric itself and be linked, for example, into the network spectrum X. And that we could affect change across the processors, the system, the fabric, the libraries, the algorithm. All of that was done simultaneously. Without CUDA to do that, I wouldn't even know where to start. My sponsor, Crusoe, was among the first clouds to offer NVIDIA's black well and black well ultra-platforms. And they just announced their NVIDIA Vera room in deployment, scheduled for later this year. But access to state of the art hardware is only part of the story. For example, most infresents already do KVCaching for a single user's forward passes. But Crusoe visited across users and GPUs. So if a thousand agents are running on the same system prompt, Crusoe only has to compute the KVCache once for it to become available to every single GPU in the cluster. This is especially important systems to get more agentech and require much longer prefixes in order to use tools and access files. In a recent benchmark, Crusoe was able to deliver up to 10 times faster, time to first token, and up to five times better throughput than VLL1. This is just one among many reasons that you should run your inference workload with Crusoe. And if you need GPUs for training, you don't need to switch clouds. Crusoe's got you covered there too. Go to cruso.ai/thorcage to learn more. So this gets to the interesting question about NVIDIA's clientele, where if 60% of your revenue is coming from these big five hyperscalers. In a different era with different customers, let's say it's professors who are running experiments. And they are helped a bunch by they need CUDA. They can't use another accelerator. They need to just run PyTorch with CUDA and have everything optimized. But if you got these hyperscalers, they have the resources to write their own kernels. In fact, they have to get that extra last 5% that they need for their specific architecture. And then, throughout back Google, mostly running their own accelerators or running GPUs and training them. But even opening eye using GPUs has Triton, which we need our own kernels. So they've down to CUDA C++. They've sort of using CUBLES and Nickel and everything. They've got their own stack, which compiles some other accelerators as well. And so if most of your customers can and do make replacements for CUDA, to what extent is CUDA really the thing that is going to make frontier AI happen on Nvidia? CUDA is a rich ecosystem. And so if you want to build on any computer first, building on CUDA first is incredibly smart. And because the ecosystem is so rich, we support every framework. If you want to create custom kernels, if you need, for example, weak contribute enormously to Triton. And so the back end of Triton, huge amounts of Nvidia technology were delighted to help every framework become as great as it can be. And there's lots and lots of frameworks. There's Triton, there's BLM, there's S-G-Lang, and there's more. And now there's a whole bunch of new reinforcement learning frameworks coming out. You got Verrol, you got Nemo-RL, you got a whole bunch of new. And then now with post-training and reinforcement learning, that entire area is just exploding. And so if you want to build on an architecture, building on CUDA makes the most sense. Because you know that the ecosystem is great. You know that if something happens, it's more likely in your code and not in the mountain of code underneath. Don't forget the amount of code that you're dealing with when you're building these systems. And something doesn't work, was it you or was it the computer? You would like it always to be you and to be able to trust the computer. And obviously we still have lots and lots and lots and lots of bugs ourselves, but our system is so well wrung out that you could at least build on top of the foundation. So that's number one, is that the richness of the ecosystem, the programmability of it, the capability of it. The second thing is if you were a developer and you were building anything at all, that single most important thing you want more than anything is install base. You want the software that you run to run on a whole bunch of the computers. You don't want to build a software, you're not building software just for yourself. You're building software for your fleet or for everybody else's fleet because you're a framework builder. And VHS CUDA ecosystem is ultimately, it's great treasure. We are now, I don't know how many several hundred million GPUs every cloud has it, goes back to A10, A100, H100, H200, you know, the L series, the P series. I mean, there's a whole bunch of them. And they're in all kinds of sizes and shapes. And if your robotics company you want that CUDA stack to actually run in the CUDA and the robot itself, we're literally everywhere. And so the install base says that once you develop the software, once you develop the model, it's going to be useful everywhere. And so the install base is just too incredibly valuable. And then lastly, the fact that we're in every single cloud makes us genuinely unique because you're an AI company and you're an AI developer, you're not exactly sure which CSP you're going to partner with and where you would like to run it and we'd run it everywhere, including on-prem for you if you like. And so I think that the richness of the ecosystem, the expansiveness of the install base and the versatility of where we are, that combination is makes CUDA invaluable. That makes a lot of sense. I guess the thing I'm curious about is whether those advantages matter a lot to your main customers. Like there's many people who they might matter for, the kind of person who can actually build their own software stack who will make up most of your revenue. Especially if you go to a world where AI is getting especially good at the things which have tight verification loops, where you can RL on them. And then this question of how do you write a kernel that does attention or MLP the most efficiently across a scale up? It's a very verifiable sort of feedback loop. And so everybody can all the hypersaylers write these custom kernels for themselves. And they might still, in videos, as. It still has great price performance, so they might still prefer to use Nvidia. But then the question is, does it just become a question of who is offering the best specs, the best flops and memory and memory bandwidth for a given dollar, where historically, Nvidia has just had, and still has the best margins and all of AI across hardware and software, 70% plus because of this CUDA mode. And the question is, "Oh, can you sustain those margins if for most of your customers, they can actually afford to build, "build instead of the CUDA mode?" - The number of engineers we have assigned to these AI labs is insane, working with them, optimizing their stack. And the reason for that is because, because nobody knows our architecture better than we do. And these architectures are not as general purpose as a CPU. The reason why a CPU is so, a CPU is kind of like a Cadillac, but it's just always, it's a nice cruiser. It never goes too fast. Everybody drives it pretty well. It's got cruise control and everything is easy. But in a lot of ways, Nvidia's GPUs are accelerators are kind of like F1 racers. And yeah, I could imagine everybody's able to drive it at 100 miles an hour, but it takes quite a bit of expertise to be able to push it to a limit. And we use a ton of AI to create the kernels that we have. And I'm pretty sure we're gonna still be needed for quite some time. And so our expertise helps our AI labs partners get another 2X out of their stack easily, oftentimes. It's not unusual that we, you know, by the time that we're done optimizing their stack or optimizing a particular kernel, their model sped up by 3X, 2X, 50%. That's a huge number, especially when you're talking about the install base of the fleet that they have, of all the hoppers and black walls that they have, when you increase it by a factor of two, that doubles the revenues. That directly translates to revenues. And Vitya's computing stack is the best performance for TCO and the world bar none. Nobody can demonstrate to me that any single platform in the world today has better performance TCO ratio, not one company. And in fact, in fact, the benchmarks are out there, Dylan's inference max is sitting out there for everybody to use. And not one TPU won't come, training won't come. I encourage them to use inference max and demonstrate their incredible inference cost. It's really, really hard. Not nobody wants to show up, ML Perf, I would welcome training to demonstrate their 40% that they claim all the time. I would love to hear them demonstrate the cost advantage of TPUs. It makes no sense in my mind. It makes absolutely zero sense on first principles and makes no sense. And so I think the reason why we're so successful is simply because our TCO is so great. There's a second 60% of our customers are the top five. But most of that business is external. For example, most of AWS's, most of Vitya and AWS's for external customers, not internal use. Most of our customers at Azure obviously, all of our customers are external, all of our customers at OCI are external, not internal use. The reason why they favor us is because our reach is so great, we can bring them all of the great customers in the world. They're all built on in Vitya. And the reason why all these companies are built on in Vitya is because our reach and our versatility is so great. And so I think the flywheel is really install base, the programmability of our architecture, the richness of our ecosystem, and the fact that there's so many AI companies in the world, there's tens of thousands of them now. And if you were one of those AI startups, what architecture would you choose? You would choose an architecture that's most abundant where the most abundant in the world. The one has the largest install base, where the most largest install base and one that has a reach ecosystem. And so that's the flywheel that's the reason why between the combination of one, our perf per dollar is so great that they have the lowest cost tokens. Second, our perf per watt is the highest in the world. And so if one of these companies, if our partners built a one gigawatt data center, that one gigawatt data center better deliver the maximum amount of revenues that and number of tokens, which directly translates to revenues, you wanted to generate as many tokens as possible, maximize the revenues for that data center, we are the highest tokens per watt architecture in the world. And then lastly, your goal is to rent the infrastructure. We have the most customers in the world. And so that's the reason why the flywheel works. Interesting. I guess the question comes down to what is the actual market structure here? Because even if there's other companies, there could be haven't been a world where there's tens of thousands of AI companies that have roughly equal share of compute. But even through these five hyperscalers, really the people on Amazon using the computer and throughout the eco-open AI and these big foundation labs who can themselves afford and have the ability to make different accelerators work. >> No, I think your assumption is premise is wrong. >> Maybe. >> Yeah. >> Let me ask you to tell you a different question. >> Come back and make me correct your premise. >> Okay. Let me just ask you a different question, which is, okay, everything you're saying. >> I still make sure that make me come back and fix because it's just too important to AI. It's too important to the future of science, it's too important to the future of the industry. That premise, the premise, look. >> Let me just first question and then I'll address it together. >> Yeah. So what do you think, if all these things are true about price performance and performance reward, etc. true, why do you think it is the case that say, Anthropic, for example, just announced a couple days ago, they have a multi-gigawatt deal with Broadcom and Google for TPUs and majority of their compute. Obviously for Google, it's TPUs and majority of their computer. So if I look at these big AI companies, it seems like a lot of their, there are some point where there's all in video. And now it's not. And so I'm curious how to square if these things are true on paper, why are they going with other accelerators? >> Yeah. Anthropic isn't a unique instance and not a trend. Without Anthropic, why would there be any TPU growth at all? It's 100% Anthropic. Without Anthropic, why would there be any training growth at all? It's 100% Anthropic. And I think that's fairly well known and well understood. It's not that there's an abundance of ASIC opportunities. There's only one Anthropic. >> But OpenEI's deals with AMD, they're building their own Titan accelerator. >> Yeah, but they're mostly, I think we could all acknowledge their vastly in video. And we're gonna still do a lot of work together. >> Yeah. >> And we're not, I'm not offended by other people using something else and trying things. If they don't try these other things, how would they know how good ours is? And sometimes you gotta be reminded of it. And we have to continuously earn the position that we're in. But they're always big claims and look at the number of ASICs that have been canceled. Just because you're gonna build a NASA, you still have to build something better than Nvidia. And it's not that easy building something better than Nvidia. It's not sensible, actually. You know, Nvidia's gotta be missing something seriously. You know, and because our scale, our velocity, we're the only company in the world that's cranking it out every single year, big leaps every single year. >> I guess their logic is that, hey, it doesn't need to be better. It just needs to be not more than 70% worse. 'Cause they're paying you 70% margins. >> No, no, no, don't forget. Even in ASICs margins really quite high. And Vias margins 70%, let's say, but in ASIC margins 65, what are you really saving? >> Oh, you mean from bought commerce? >> Yeah, sure. You gotta pay somebody. And so I think the ASIC margins are incredibly good from what I can tell. And they believe it's so too. And so they're quite proud of their incredible ASIC margins. And so you asked the question, why a long time ago, we just didn't have the ability to do it. And this is, and at the time, at the time I didn't deeply internalize how difficult it would be to build a foundation, AI lab, like OpenAI and Anthropic. And the fact that they needed huge investments from the supplier themselves. We just weren't in a position to make the multi-billion dollar investment into Anthropic so that they could use our compute. But Google and AWS were. And they put in huge investments in the beginning so that Anthropic in return used their compute. We just weren't in a position. to do so at the time. Nor did I, I would say my mistake is I didn't deeply internalize that they really had no other options. That a VC would never put in $5, $10 billion investment into an AI lab with the hopes of it turning out to be unthropic. And so that was my miss. But even if I understood it, I don't think we would have been in a position to do that at the time. But I'm not going to make that same mistake again. And I'm delighted to invest in OpenAI and I'm delighted to help them scale. And I believe it's essential to do so. And then when I was able to, anthropocame to us, I'm delighted to be an investor, delighted to help them scale. But we just weren't at the time able to do so. If I could rewind everything, Nvidia could have been as big back then as we are now, I would have been more than happy to do it. This is actually quite interesting, which is for many years, Nvidia has been this decompanied in AI, making lots of money. And now you're investing it. It's been reported that you've done up to 30 billion in OpenAI and 10 billion in Anthropic. But now, their valuations have increased. And I'm sure they'll continue to increase. And so, overall, these many years, you were giving them the compute. You saw where I was headed. And then they were worth like one tenth what they are now a couple of years ago or even a year ago, in some cases. And you had all this cash. There's a world where either Nvidia themselves becomes a foundation lab that does a huge investment to make that possible or has made the deals you've made now occurred valuations much earlier on. And you had the cash to do it. So I am curious, actually, why not have done it earlier? We did it as soon as we could have. And if I could have, I would have done it even earlier. At the time that Anthropic needed us to do it, we just weren't in a position to do it. It wasn't in our sensibility to do so. How so, like a cash thing? Or just the level of investment. We never invested outside the company at the time. And not that much. And we didn't realize we needed to. I always thought that they could just go raise VCs for God's sake, like all companies do. But what they were trying to do couldn't have been done through VCs. What OpenAI wanted to do couldn't have been done through VCs. And I recognized that now. I didn't know what then. But that's their genius. That's why they're smart. So they realized that then that they had to do something like that. And I'm delighted that they did. And even though, even though we caused Anthropic to have to go to somebody else, I'm still happy that it happened. Anthropics existence is great for the world. I'm delighted for it. I guess you still are making a ton of money. And we're making way more money. Quarter after quarter. It's still okay to have regrets. So the question still arises. Okay, well, now that we're here, you have all this money that you keep making. What should I be doing with it? And there's one answer which says, look, there's this whole middleman ecosystem that has popped up for converting CAPEX into OPEX for these labs so that they can rent compute. Because the ships are really expensive. They make a lot of money over their lifetime because the AI model is getting better. The value that they generate their tokens is increasing. But they're expensive to set up. Nvidia has the money to do the CAPEX. So, and in fact, you are, it's been reported your backstopping core. We have up to 6.3 billion and have invested 2p. But yeah, why doesn't Nvidia become a cloud themselves? Why doesn't it become a hyper-scholar themselves? And when this compute out, you have all this cash to do it. This is a philosophy of the company and I think is wise. We should do as much as needed as little as possible. And what that means is the work that we do with building our computing platform, if we don't do it, I genuinely believe it doesn't get done. If we didn't take the risk that we take, if we didn't build NVLink the way we built, if we didn't build the whole stack, if we didn't create the ecosystem the way we did it, if we didn't dedicate ourselves to 20 years of CUDA while losing money most of that time, if we didn't do it, nobody else would have done it. If we didn't create all the CUDA X libraries so that they're all domain-specific, you know, this several decade and a half ago, we pushed into domain-specific libraries because we realized that if we didn't create these domain specific libraries, whether it's for ray tracing or image generation or even the early works of AI, these models, if we didn't create them for data processing, structure data processing, or vector data processing, if we didn't create them, nobody would. And I am completely certain of that. We created a library for computational lithography called Kulitho, if we didn't create it, nobody would have. And so accelerated computing went advanced the way it has if we didn't do what we did. And so we should do that. We should dedicate our company all of our might wholeheartedly to go do that. However, the world has lots of clouds. If I didn't do it, somebody show up. And so following the recipe, the philosophy of doing as much as needed, but as little as possible, as little as possible, that philosophy exists in our company today. And everything I do, I do it with that lens. In the case of clouds, if we didn't support CoreWeave to exist, these Neo Clouds, these AI clouds, when exist, if we didn't help CoreWeave exist, they would not exist. If we didn't support end scale, they wouldn't be where they are today. If we didn't support Nebius, they wouldn't be what they are today. Now they are, they're doing fantastically. Is that a business model? No. And so we're trying to, we invest in our ecosystem because I want our ecosystem to thrive. And I want our, I want, I want the architecture and I want AI to be able to connect with as many industries as possible, as many countries as possible, and make it possible for, you know, the planet to be built on AI and to be built on the American tech stack. And so, so though that vision, I think, is exactly what we're pursuing. Now one of the things that that you mentioned, there are so many great amazing foundation model companies and we try to invest in all of them. And this is, this is another thing that we do. We don't pick winners. And we like, we, we need to support everyone. And it's part of our, part of our, our joy of doing so. It's, it's an imperative to our business. But we also go out of our way not to pick winners. And so when I, when I invest in one of them, I invest in all of them. What, why do you go out of your way to not pick winners? Because it's not our job to. Number one, number two, when NVIDIA first started, there were 60 graphics companies, 63 D graphics companies. We are the only one that survived. If you would have taken those 60 companies, 60 graphics companies and ask yourself which one was going to make it. NVIDIA would be the top of that list not to make it. You know, this is long before you. But NVIDIA's graphics architecture was precisely wrong. It's not a little bit wrong. We created an architecture that was precisely wrong. And, and it was an impossible thing for developers to support. It was never going to make it. We reasoned about it for good, for, from good first principles. But we ended up in the wrong solution. And, and everybody would kind of, everybody would have counted us out. And, and here we are. And so I'm, I'm, I have enough humility to recognize that, you know, don't, don't pick winners. Yeah. Either let them all take care of themselves or take care of all of them. What one thing I didn't understand is, you said, look, we're not prioritizing these new clouds. Just because there are new clouds and we want to prop them up. But you also said, you listed a bunch of new clouds and you said they wouldn't exist if it wasn't for NVIDIA. Yeah. And so how are those two things compatible? First of all, they, they need to want to exist. And they come to ask us for help. And when they, when they, when they want to exist and have, they have a business plan and they, you know, they have expertise and, you know, they have the passion for it. They obviously have to have some capability, some selves. But if at the end of the day, they need some investment and we're together off the ground, we, we would be there for them. But, but the sooner they get their flywheel going, you know, your question was, do we want to be in the financing business? The answer's no. Yeah. We don't want to be, we want to, we, because there are people in the financing business. And we rather work with all of the people who are in the financing business than to be a financier ourselves. And so, so I think the, the, our goal is to focus on what we do, keep our business model as simple as possible, support our ecosystem. When someone like, like, open AI needs an investment of $30 billion scale, because it's still before their IPO. And, and we deeply believe in them. We deeply believe that I deeply believe that that they're going to be, they're going to be and well they're an extraordinary company already. today, they're going to be an incredible company. The world needs them to exist. The world wants them to exist. They have everything, they have the win of their back. Let's support them and let them scale. So those investments will do because they need us to do it. But we're not trying to do as much as possible. We're trying to do as little as possible. I spend a way too much time copy-pasting tax back and forth from Google Docs to Chat Pots. And so I built what's basically a cursor for writing, which operates the way I think an AI co-researcher should operate. I can tag it and it can talk with me through inline comment threads and help me dig deeper and brainstorm. I brought this entire thing over the weekend with cursor and their new composer to model. With a lot of agenda coding tools, I feel like I have no idea what's going on under the surface. I just have to relinquish control and hope for the best. But cursor, let me try a bunch of different ideas while staying on top of the implementation. I did most of my brainstorming in the agent's window. And after I got some basic files in place, I used a different window to track changes. The few times that I needed to make a quick tweak by hand, I just used the editor. If you want to try my AI co-researcher yourself, I have linked the GitHub repo in the description. And if you have a tool that you've been wanting to build, you should make it happen. Go to cursor.com/sworkache to get started. This may be sort of an obvious question, but we've lived many years in this situation where there's a shortage of GPUs. It's grown now because models are getting better. We have a shortage of GPUs. Yes. And Nvidia is known for giving up the scarce allocation, not just based on high spitter, but rather on, hey, we want to make sure that these new clouds exist. Let's give some to Corby. Let's give some to Crusoe. Let's give some to Lambda. Why is it good for Nvidia? First of all, would you agree with this characterization of capturing the market? No, no. Your premise is just wrong. Yeah. We're sufficiently mindful about these things. We're very mindful about these things. First of all, if you don't place a PO, all the talking in the world won't make a difference. And so until we get a PO, what are we going to do? And so the first thing is we work really hard with everybody to get a forecast done. Because these things take a long time to build. And the data centers take a long time to build. And so we align ourselves with demand and supply and things like that through forecasting. As job, job number one, number two, everybody who, you know, we've tried to forecast with it was with as many people as possible, but in the final analysis, you still have to place an order. And maybe, maybe for whatever reason, you didn't place your order. What can I do? And so at some point, first and first out, but beyond that, if you're not ready, because your data center's not ready, or serving components aren't ready to enable you to stand up a data center, we might decide to serve another customer first. That's just maximizing the throughput of our, of our, our own factory. And so we might do some adjustments there. Aside from that, the prioritization is first and first out. Yeah, you got, you got to place a PO, even on place a PO. Now, of course, there are stories about that, you know, like, for example, all of this kind of started from, from, it was an article about Larry and Elon having dinner with me where they, they begged for GPUs. That never happened. We, we absolutely had dinner. And we absolutely had dinner. And it was, it was a wonderful dinner. And no time did they beg for GPUs. And so it, it, they just had the place in order. And once they placed no order, we do our best to get the capacity to them. Yeah. We're not complicated. Okay. So it sounds like there's a queue. And then based on whether your data center is ready and when you put, place a purchase order, you get them a certain time. But it still doesn't sound like high spit or just gets it. Is there a reason to do it? We never do that. Okay. We never buy an eye just to high spit or because it's, it's a bad business practice. You, you set your price, you set your price and then, and then people decided buy it or not. And, and, um, they're, they're, I, I understand that, that others in the chip industry, um, uh, change their prices when demand is higher. But we just don't, we just don't, that's just never been a practice of ours. You can count on us. You know, I, I prefer to be, to be, um, uh, dependable, uh, to be the foundation of the industry. And I, I, you don't need, you don't need the second guess. You know, if, if you, if I quoted you a price, um, we quoted you a price. That's it. And if demand goes through the roof, so be it. And I'm the other end. That's why you have a collective relationship with TSMC, right? Yeah. Yeah. Uh, and Vity has been in business, we've been doing business with them for, uh, I guess coming up on 30 years. And, and Vity and TSMC don't have a legal contract. There's, there is always some rough justice and, um, sometimes I'm right, sometimes I'm wrong. Uh, sometimes I got, I got a better deal. Sometimes I got a worse deal. Uh, but overall in the, in the whole, the relationship is incredible. And, and I can completely trust them. I can completely depend on them. And, and our, our, one of the things that we, you can count on within Vitya is that next year, this year, Vera Rubin is going to be incredible. Next year, Vera Rubin ultra will come the year after that Feynman will come and the year after that, I haven't introduced the name yet. And so, so every single year you can count on us. And this is an, you, you're going to have to go find another ASEC team in the world. Pick your ASEC team where you can say, I can bet the farm of, I can bet my entire business that you will be here for me every single year, your cost, your token cost will decrease by an order of magnitude every single year. I can count on it, like I can count on the clock. Well, I just said something about TSMC. No other foundry in history. Can you possibly say that? You can say that by Vitya today. You can count on us every single year. If you would like to buy a billion dollars with a AI factory compute, no problem. If you like to buy a hundred million dollars, no problem, you'd like to buy $10 million dollars or just one rack, not a problem or just one graphics card. Okay, no problem. If you would like to place an order for a hundred billion dollar AI factory, no problem. We're the only company in the world where you can say that today. I can say that about TSMC as well. I want to buy one, buy one billion, no problem. We just got to go through the process of planning for it and all the things that mature people do. I think this ability for Vitya to be the foundation of the world's AI industry, this is a position that has taken several decades to arrive at, enormous commitment and enormous dedication and the stability of our company, the consistency of our company is really, really important. I want to ask you about China. Yeah. I always like to take, I don't actually don't know what I think about whether it's good to sell chips China or not, but I've played levels out against my guess. When Darryl was on who supports X-Work control, they asked him, "Why can't America and China both have country geniuses in the data center?" But since you're on the opposite side, I'll ask you in the opposite way. When I think about it as Anthropic announced a couple days ago, this model myth was not even releasing publicly because they say it has such cyber offensive capabilities that we don't think the world is ready until we make sure these year-ades are patched up. But they say it found thousands of high-civility vulnerabilities across every major operating system, every browser. It found one in OpenBSD, which is this operating system that's been specifically designed to not have the year-days and found one for 27 years that's existed. If Chinese companies and Chinese labs and Chinese government had access to the AI chips to train a model like Cloud Mythos with these cyber offensive capabilities and run millions of instances of it with more compute, the question is, "Oh, is that a threat to American companies to American national security?" First of all, Mythos was trained on fairly mundane capacity and a fairly mundane amount of it. By an extraordinary company. The amount of capacity and the type of compute that's it was trained on is abundantly available in China. You just have to first realize that chips exist in China. They manufacture 60 percent of the world's mainstream chips, maybe more. It's a very large industry for them. They have some of the world's greatest computer scientists. As you know, most of the AI researchers in all of these AI labs, most of them are Chinese. They have 50 percent of the world's AI researchers. And so the question is, if you're concerned about them, what is the considering all the assets they already have? They have an abundance of energy. They have plenty of chips. They got most of the AI researchers. If you're worried about them, what is the best way to create a safe world. Well, victimizing them, turning them into an enemy likely isn't the best answer. They are an adversary. We want the United States to win. But I think having a dialogue and having research dialogue is probably the safest thing to do. This is an area that is glaringly missing because of our current attitude about China as an adversary. It is essential that our AI researchers and their AI researchers are actually talking. It is essential that we try to both agree on how to what not to use the AI for. With respect to finding bugs in software, of course, that's what AI is supposed to do. Is it going to find bugs in a lot of software? Of course, there are lots and lots of bugs. There are lots of bugs in the AI software. That is what AI is supposed to do. I am delighted that AI has reached a level where it could help us be so much more productive. One of the things that is under emphasized is the richness of ecosystem around cybersecurity, AI cybersecurity, and AI privacy, and AI safety. That whole ecosystem of AI startups that are trying to create this future for us, where you have one AI agent that's incredible, surrounded by thousands of AI agents keeping it safe, keeping it secure. That future surely is going to happen. The idea that you're going to have an AI agent running around with nobody watching after it is kind of insane. We know very well that this ecosystem needs to thrive. It turns out this ecosystem needs open source. This ecosystem needs open models. They need open stacks so that all of these AI researchers and all these great computer scientists can go build AI systems that are as formidable and can keep AI safe. One of the things that we need to make sure that we do is we keep the open source ecosystem vibrant. That can't be ignored. A lot of that is coming out of China. We have to not suffocate that. With respect to China, we want to have, of course, want United States staff as much computing as possible. We're limited by energy, but we've got a lot of people working on that, and we have to not make energy a bottleneck for our country. But what we also want is we want to make sure that all the AI developers in the world are developing on the American tech stack and making the contributions, the advancements of AI, especially when it's open source available to the American ecosystem. It would be extremely foolish to create two ecosystems. The open source ecosystem and it only runs on the Chinese tech, a foreign tech stack and a closed ecosystem. That runs on the American tech stack. I think that would be a horrible outcome for United States. Since there are a lot of things, let me just triage the response. I think the concern going back to the flop difference in the hacking is, yes, they have compute, but there's some estimates that because they're at 7 nanometer, they don't have UVs because of chip making expert controls, the amount of flops that you're able to actually produce, they have like one tenth the amount of flops that the US has. With that, could they train eventually a model like Mythos? Yes, but the question is because we have more flops, American labs are able to get to these double capabilities first, because in Thropigot 2.1, they say, okay, we're going to hold on to it for a month while all these American companies give them access to it, they're going to patch up all their vulnerabilities, and now we release it. Furthermore, even if they train a model like this, the ability to deploy that scale, if you had a cyber hacker, it's much more dangerous if they have a million of them versus a thousand of them, so that inference compute really matters a lot. In fact, the fact that they have so many researchers are so good is the thing that makes it so scary because what is it that makes those engineers researchers more productive is compute. If you talk to any AI lab in America, they say the thing that's bottleneck and they miscompute. They are quotes from deep seek founder or coin leadership or whatever, they say the thing we're bottleneck on is compute. So then the question is, isn't it better that we get to get American companies because they have more compute, get to get to the level of spot or mythos, level capabilities first. Prepare our society for it before China can get to it because they have less compute. We should always be first and we should always have more. But in order for that outcome for you to, will you describe to be true, you have to take it to the extremes. They have to have no compute. And if they have some compute, the question is how much is needed. The amount of compute they have in China is enormous. You're talking about a country. There's a second largest computing market in the world. If they want to deploy aggregate, they're compute. They got plenty of compute to aggregate. But is that true? I mean, there's people do these estimates and they're like, well, this make is actually behind on the process. No, it's a they're actually just. I'm about to tell you. The amount of energy they have is incredible. Isn't that right? AI is a parallel computing problem, isn't it? Why can't they just put four, ten times as much chips together because energy is free? They have so much energy. They have data centers that are sitting completely empty, fully powered. They've, you know, they have ghost cities. They have ghosts, ghost data centers. They have so much capacity of infrastructure. If they wanted to, they just gang up more chips, even if they're seven nanometer and their capacity of building chips is one of the largest in the world. The semiconductor industry knows that they monopolize mainstream chips. They over capacity. They have too much capacity. And so the idea that China won't be able to have AI chips is completely nonsense. Now, of course, if you ask me, would the United States be further ahead if the entire world had no compute at all? But that's just not an outcome. That's not a scenario that's true. They have plenty of compute already. The amount of threshold they need for the concern you're worried about, they've already reached that threshold and beyond. And so, I think you misunderstand that AI is a five-layer cake. And at the lowest layer, layer is energy. When you have abundant of energy, it makes up for chips. If you have abundance of chips, it makes up for energy. For example, United States is scarce on energy, which is the reason why Nvidia has to keep advancing our architecture and do this extreme co-design so that with the few chips that we ship, with the few chips because the amount of energy is so limited, our throughput per watt is off the charts. But if your amount of watts is completely abundant, it's free. What do you care about performance for a watt for? You can use old chips to do so. So seven nanometer chips are essentially hopper. The ability for hopper, I gotta tell you, today's models are largely trained on hopper. Hopper generation. So hopper, seven nanometer chips are plainly good. The abundance of energy is their advantage. But then there's a question of, okay, well, can they actually manufacture enough chips given their-- But they do. What's the evidence? Huawei just had the largest single year in the history of their company. How many chips did they ship? A ton. Millions. Millions is way more than anthropic house. So there's a question of how much logic, Smith and Chef, then there's a question of how much memory-- I'm telling you what it is. They have plenty of logic and they have plenty of HBM2 memory. Right. But as you know, the bottleneck often in training and doing inference on these models is the amount of bandwidth. So if you HBM2, I don't know the numbers offhand, but like versus the newest thing you have, you know, you can be almost in order of magnitude difference of memory bandwidth, which is-- Huawei's a networking company. Huawei's a networking company. But that doesn't change the fact that you need a UV for the most advanced HBM. Not true. Not at all true. You could gang them together, just like we gang them together with NBLink72. They've already demonstrated silicon photonics sub-connecting all of these compute together into one giant supercomputer. The fact that it matters, their AI development is going just fine. And the best AI researchers in the world, because they are limited in compute, they also come up with extremely smart algorithms. Remember, I just-- what I said, I said that Moore's Law is advancing about 25% per year. However, through great computer science, we could still improve algorithm performance by 10X. What I'm saying is great computer science is where the lever is. There is There's no question. MOE is a great invention. There's no question all the incredible attention mechanisms reduce the amount of compute. We have got to acknowledge that most of the advances in AI came out of algorithm advances, not just the raw hardware. Now, if most of the advances came from algorithms and computer science and programming, tell me that their army of AI researchers is not their fundamental advantage. And we see it. DeepSeek is not in consequential advance. And the day that DeepSeek comes out on Huawei first, that is a horrible outcome for our nation. - Why is that? Because I mean, currently you can have a model like DeepSeek that can run on any accelerator, if it's open source. Why would that stop being the case in the future? - Well, suppose it doesn't. Suppose it's optimized for Huawei. Suppose it's optimized for their architecture. It would put others at a disadvantage. You describe the situation that I perceived, I perceived to be good news, that a company developed software, developed an AI model, and it runs best on the American tech stack. I saw that as good news. You set it up as a premise that it was bad news. I'm gonna give you the bad news, that AI models around the world are developed and they run best on not American hardware. That is bad news for us. - I guess I just don't see the evidence that there's these huge disparities that would prevent you from switching accelerators. There's American labs, running their models across all the clouds, across all the different cylinders. - You take a model that's optimized for Nvidia and you try to run on something else. - But the American labs do that. - And they don't run better. Nvidia's success is perfect evidence. The fact that AI models are created on our stack runs best on our stack. How is that illogical to understand? - I'm just looking, look, and THROPX models are run on GPUs, they're run on Trinium, they're run on GPUs. - A lot of work has to go into it to change. But go to the global south, go to the Middle East, coming out of the box, if all of the AI models run best on somebody else's tech stack, you've got to be arguing some ridiculous claim right now that that's a good thing for our United States. - But I guess I don't understand arguments. So like if, say, Chinese companies get to the next mythos first, they find that all the security of an American software first. But they can do it on Nvidia hardware and they ship it to the global south, they doesn't even read the hardware. Like how is that good? I mean, okay, the ones that are gonna be hard to read. - It's not good. - It's not good. - It's not good, so let's not let it happen. - Why do you think it's perfectly fungible? That if you didn't ship them computer with exactly your place of highway, they are behind, right? They have where it shifts in you. - It's completely, there's evidence right now. Their chip industry is gigantic. - You can just look at the flop or bandwidth or memory comparisons between the H200 and the Huawei 910C. It's like half half a third. - They use more of it. They use twice as many. - I guess it seems like argument is they have all this energy that's ready to go, right? And they need to fill it with chips. - And they're good at manufacturing. - And I'm sure eventually they would be able to just, I'll manufacture everybody, but there's these few critical years. - What is the critical year you're talking about? - These next few years, we've got these models that are gonna do all the cyber attacks. - If the critical years, the next critical years is critical, then we have to make sure that all of the world's AI models are built on American tech stack. These critical years. - Okay, how would that prevent, if they're built on American tech stack, how would that prevent them from, if they have more advanced capabilities from launching the mythos equivalent cyber attacks on their source? - They're not guarantee you the way. But if you have it earlier, we're gonna prepare for it. - Listen, why are you causing one layer of the AI industry to lose an entire market so that you could benefit another layer of the AI industry. There's five layers, and every single layer has to succeed. The layer that has to succeed most is actually AI applications. Why are you so fixated on that AI model, that one company, for what reason? - Because those models make possible, these incredibly offensive capabilities, and you need computer on them. - The energy, the chips, the ecosystem of AI researchers make it possible. - A few months ago, Jane Streets spent about 20,000 GPU hours trading backdoors into three different language models. Then, they challenged my audience to find the trigger phrases. I just kind of with Rickson, who designed the puzzle about some of the solutions that Jane Streets received. - If you think the base model was here and the backdoor model was here, you can kind of linearly interpolate the weights to like adjust the strength of the backdoor, but you can also extrapolate it to make the backdoor even stronger. And in some cases, if you make it strong enough, the model will just regurgitate what the response phrase was supposed to be. - So if you keep amplifying the difference between the base version and the backdoor version, eventually it should spread out the trigger phrase. But this technique only worked on two out of the three models. Even Rickson isn't sure why he didn't work on the other. Being able to verify that a model only does what you think it does is one of the most important open questions in AI security. If this is the kind of problem that excites you, Jane Streets is hiring researchers and engineers. Go to janestreets.com/thorkash to learn more. Okay, stepping back, it has to be the case that Chionis able to build enough seven nanometer capacity. And remember, there's still stuck on seven nanometer. Well, you will move on to three nanometer and then two nanometer or 1.6 nanometer with Feynman. So while you're on 1.6 nanometer, there's still gonna be on seven nanometer. And they have to produce enough of it to make up for the shortfall. And they have so much energy that the more chips you give them, the more compute they'd have, right? Like, so there's comes out as a question of, ultimately they are getting more compute, compute is in input to training and inference. I just think you speak in absolutes. I think that United States ought to be ahead. The amount of compute in United States is a hundred times more than anywhere else in the world. The United States ought to be ahead. Okay, the United States is ahead. In video builds, the most advanced technologies, we make sure that the US labs are to first to hear about it and the first chance to buy it. And if they don't have enough money, we even invest in them. We wanna do everything we can to make sure that the United States is ahead. Number one point, do you agree? And we're doing everything we can to do that. But how is shipping chips to China? Keeping the US and China? If they're bottleneckin, we have very, we got very, very Rubin for United States. We have very Rubin for United States. Now United States, am I in United States? Do you consider me partnering United States? Yes. In video. You consider Nvidia a United States company. Okay, number one. Why is it that we don't come up with a regulation? That's more balanced so that Nvidia can win around the world instead of giving up the world. Why would you want United States to give up the world? The chip industry is part of the American ecosystem. It's part of American technology leadership. It's part of the AI ecosystem. It's part of AI leadership. Why is it that your policy, your philosophy leads to United States giving up a vast part of the world's market? The claim here is, I'll for it, Darryl had this quote where he said, "It's like Boeing bragging that we're selling "nor through a new expo the missile casings "are made by Boeing." And that's somehow enabling the US technology stack. Like fundamentally, you're giving them this capability. I'm carrying AI to anything that you just mentioned is lunacy. But AI similar to Enrich uranium, right? And then it can have positive uses, it can have negative uses. We still don't want to send Enrich uranium to other countries. Who's sending Enrich, the analogy is Enrich uranium is like, "It's a lousy analogy. "It's an illogical analogy." But if it's, if that compute can run a model, that can do zero-day exploits against all Americans off-ware, how is that not a weapon? First of all, the way to solve that problem is to have dialogues with the researchers, dialogues with China and dialogues with all the countries to make sure that people don't use technology in that way. That's a dialogue that has to happen, okay? Number one, number two, we also need to make sure that the United States is ahead. Everything that Ruben, Vera Ruben, Blackwell is available in the United States in abundance. Mounds of it, obviously, are results with show-it. Abundance, a tons of it, tons of it. The amount of computing we have is great. We have amazing AI researchers here. It's great. We have to stay ahead. However, we also have to recognize that AI is not just a model, that AI is a five-year-layer cake, that AI industry matters across every single layer. And we want United States to win at every single layer, including the chip layer, and conceding the entire market. It's not gonna allow United States to win the technology race, long-term in the chip layer, in the computing stack. That is just a fact. I guess then the crux comes down to, how does selling them chips now help us win in the long-term? Like Tesla sold extremely good electric vehicles to China for a long time. iPhone's are sold in China, extremely good. They didn't cost them lock-in. China will still make their version of EVs and their dominating or smart-fronts-dominating. When you started the conversation today, you would acknowledge it. You acknowledged that NVIDIA's position is very different. Use words like "mote." The single most important thing to our company is our richness of our ecosystem, which is about developers. 50% of the AI developers are in China. We don't want to, we shouldn't, the United States should not give that up. But we have a lot of NVIDIA developers in the US, and that doesn't prevent American labs from also being able to use other accelerators in the future. In fact, right now they're using other accelerators as well, which is fine and great. I don't see why that wouldn't be the case in China as well. If you sell them in video trips, just the same way that Google can use TPUs and Nvidia. - We have to keep innovating and, you know, as you probably know, our share is growing not decreasing. The premise that even if we compete in China, that we're gonna lose that market anyways, I don't, you're not talking to somebody who woke up a loser. And that loser attitude, that loser premise makes no sense to me. We are not, we're not a car. We are not a car. It, the fact that I can buy a car, this car brand one day and use another car brand another day, easy. Computing is not like that. There's a reason why the X86 deal exists. There's a reason why arm is so sticky. These ecosystems, these ecosystems are hard to replace. They cost an enormous amount of time and energy and most people don't want to do it. And so it's our job to continue to nurture that ecosystem, to keep advancing the technology, so that we could compete in the marketplace. Conceiting a marketplace based on the premise you described, I simply can acknowledge that. It makes no sense. Because I don't think the United States is a loser. Our industry is now a loser. And that losing proposition, that losing mindset, makes no sense to me. Okay, I'll move on. I just, I just want to make sure that. You don't have to move on. I'm enjoying it. Okay, great. Yeah, yeah. Yeah. I appreciate that. Yeah, sure. But I think that maybe the crocs, and thanks for walking around the circles with me, because then I think it helps bring out what the crocs here is. The crocs is you're going to extremes. Your argument starts from extremes. That if we give them any compute at all, in this narrow moment, we will lose everything. No, I think what my argument is. Those extremes, they're childish. Let me just tell you what my argument is. They're childish. Yeah. The idea is not that there is some key threshold of compute. Is that any marginal compute is helpful, right? So if you have more compute, you can train a better model. And I just want you to acknowledge that any marginal sales for American technology industry is beneficial. I actually don't, I mean, if the AI models that run on those chefs, are capable of cyber offensive capabilities, or training models are capable of cyber-independence, running more models of those instance, it is not a nuclear weapon, but it enables a weapon of a kind. The logic that you use, you might as well say to microprocessors and DRAMs. You might as well say to electricity. But in fact, we do have extra controls on the technology that is relevant to making the most advanced DRAM, right? We have all kinds of extra controls on China for all kinds of shabby things. We saw a lot of DRAM and CPUs into China. And I think it's right. I guess this is back to the fundamental question of is AI different, right? If you have the kind of technology, they can find these year days in software. Is that something where we want to minimize China's ability to get their first-- We want to be-- Do not start to be ahead. We can control that. How do we control that if the chips are already there and they're using that to train that model? We have tons of compute. We have tons of AI researchers. We're racing as fast as we can. Again, we have more nuclear weapons than anybody else, but we don't want to send in rich and uranium anywhere. We're not enriched uranium. It's a chip and it's a chip that they can make themselves. But there's a reason they're buying it from you, right? And we have quotes from the founders of Chinese companies that say they were bottling that thing. Because our chips are better. On balance, our chips are better. There's just no question about it. In the absence of our chip, can you acknowledge that Huawei had a record year? Can you acknowledge that a whole bunch of chip companies have gone public? Can you acknowledge that? Can you acknowledge that-- Can you also acknowledge that the fact that we used to have a very large share in that market and we no longer have the large share in that market, we can also acknowledge that China is about 40% of the world's technology industry. That market to leave that market, can see that market for the United States technology industry is a disservice to our country. It is a disservice to our national security. It is a disservice to our technology leadership. All for the benefit, all for the benefit of one company. It makes no sense to me. I guess I'm confused of-- it feels like you're making two different statements. One is that we're going to win this competition with Huawei because our chips are going to be way better if we're allowed to compete. And another is that they would be doing the same exact thing without us anyways. Right, how can those two things be the same true at the same time? It's obviously true. In the absence of a better choice, you'll take the only choice you have. How is that illogical? But so logical-- The reason they want-- In relationships, they're better. Better is more compute. More compute means you can train better. No. It's better because it's easier to program. We have a better ecosystem. Whatever the better is, whatever the better is. And of course, we're going to send them compute. So what? So what? The fact that a matter is we get the benefit-- don't forget, we get the benefit of American technology leadership. We get the benefit of developers working on the American tech stack. We get the benefit as those AI models diffuse out into the rest of the world. The American tech stack is therefore the best for it. We can continue to advance and diffuse American technology. That, I believe, is a positive. It's a very important part of American technology leadership. Now, the policy that you're advocating resulted in the American telecommunication industry being policy out of basically the world, to the point where we don't control our own telecommunications anymore. I don't see that as smart. It's a little narrow-minded, and it led to unintended consequences that I'm describing to you right now that you seem to have a very hard time understanding. OK. Let's just step back. It seems like the crux here is there's a potential benefit and there's a potential cost. And we're trying to figure out is the benefit worth the cost. I guess I'm trying to get you to acknowledge the potential cost that compute is an input to training powerful models. Powerful models do have powerful offensive capabilities like cyberattacks. It is a good thing that American companies got to cloud Mythos level capabilities first. And then now they're going to hold off on this capability so that the American companies and American government can make their software more protected before this level capability is announced. If China had had more computer, or had more car compute, if we could have had made a Mythos level model earlier and deployed it widely, that would have been very bad. One of the reasons that hasn't happened is that we have more compute thanks to companies like Nvidia in America. That is a cost of sending to China. And so let's leave the benefit of Cypher's second, do you acknowledge that this is a potential cost? I will also tell you the potential cost is we allow one of the most important layers of the AI stack, the chip layer, to concede an entire market. The second largest market in the world so that they could develop scale, so that they could develop their own ecosystem so that future AI models are optimized in a very different way than the American tech stack. As AI diffuses out into the rest of the world, their standards, their tech stack, will become superior to ours because their models are open. I guess I just believe enough in Nvidia's kernel engineers and CUDA engineers to think that they could optimize. AI is more than kernel optimizations, you know. Of course, but there's so many things you can do from distilling to a model that's well-fit for your chips. We're going to do our best. You have all this offer. There's a long-term lock-in. So if they have this slightly better opens or a model for a while. China is the largest contributor to open source software in the world. Fact. Right. China is the largest contributor to open models in the world. Fact. Today it's built on the American tech stack in Videos. Fact. All five layers of the tech stack for AI is important. United States ought to go win all five of them. They're all important. The one that is the most important, of course, is the AI application layer. The layer that diffuses into society, the one that uses it most, will benefit from this industrial revolution most. But my point is that every layer has to succeed. If we scare this country into thinking that AI is somehow a nuclear bomb, so that everybody hates AI and everybody's afraid of AI, I don't know how you're helping the United States. You're doing a disservice. If we scare everybody out of doing software engineering jobs because it's going to kill every software engineering job, and we don't have any software engineers as a result of that, we're doing a disservice to United States. If we scare everybody out of radiology, so nobody wants to be a radiologist because computer vision is completely free, and no AI is going to do a worse job than radiologists. And we misunderstand the difference between a job and the task, the job of a radiologist, patient care, task to read a scan. If we misunderstand that so profoundly, and we scare everybody out of doing to radiology school, we're not going to have enough radiologists and good enough health care. And so I'm making the case that when you make these make a premise that is so extreme, everything goes from zero or infinity, we end up scaring people in a way that's just not true. Life is not like that. Do we want United States to be first? Of course we do. Do we need to be a leader in every layer of that? stack. Of course we do. Of course we do. Is today you're talking about mythos because mythos is important? Sure, that's fantastic. But in a few years time, I'm making you to prediction that when we want the American tech stack, when we want American technology to be diffused around the world, out to India, out to the Middle East, out to Africa, out to Southeast Asia, when our country would like to export. Because we would like to export our technology. We would like to export our standards. On that day, I want you and I to have that same conversation again. And I will tell you exactly about today's conversation, about how your policy and how what you imagined. Literally, causing the United States to concede the second largest market in the world for no good reason at all. We shouldn't concede it. If we lose it, we lose it. But why do we concede it? Now, nobody is advocating. Nobody is advocating and all are nothing. Nobody's advocating all are nothing, meaning we ship everything to China at all times. Nobody's advocating that. We should always have the best technology here. We should always have the most technology here. And the first, but we should also try to compete and win around the world. Both of those things can simultaneously happen. It requires some amount of nuance, some amount of maturity. Instead of absolutes, the world is just not absolutes. Okay. The argument hinges on they built the models that are specified for their architect. They're the best chips that they make in a few years. And those chips get exported around the world. That sets the standard. Because of EUV, export controls, as we said, you're going to move on to 1.6 nanometer. This is going to be a 7 nanometer, even after a few years from now. And it makes sense that domestically, they would prefer, hey, we got so much energy, we can manufacture it such scale, we'll still keep using 7 nanometer. But the exporting thing, their 7 nanometerships have to be competitive against, well, you're 1.6 nanometer chips. And their models have to be so far optimized for the 7 nanometer that's better to run their models on 7 nanometer than to run their models on your 1.6 nanometer. Can we just look at the facts then? Okay. Is blackwell 50 times more advanced lithography than Hopper? Is it 50 times? Not even close. I just kept saying it over and over again, more's laws dead. Between Hopper and Blackwell, from the transistors themselves, call it 75%. It was three years apart. 75%. Blackwell is 50 times Hopper. My point is architecture matters. Computer science matters. Some encounter physics matters as well. The computer science matters. AI, the impact of AI largely comes from the computing stack, which is the reason why kuda is so effective, which is the reason why kuda is so so beloved. It's an ecosystem of computing architecture that allows for so much flexibility that if you wanted to change an architecture completely, create something like M.O.E. Create something like diffusion. Create something that's disaggregated. You could do so. It's easy to do. And so the fact that a matter is AI is about the stack above as much as it is about the architecture below. To the extent that we have architectures and software stacks that are optimized for our stack, for our ecosystem, it is obviously good because we started the conversation today about how NVIDIA's ecosystem is so rich, why people always love programming kuda first. They do. They do. And so did the researchers in China. But if we are forced to leave China, if we're forced to leave China, it would be, it would be, well, first of all, it's a policy mistake. Obviously has backlash. It has backlash. Obviously, it has far, you know, has turned out badly for the United States. It enabled it accelerated our chip industry. It forced all of their AI ecosystem to focus on their internal architectures. It's not too late, but nonetheless, it has already happened. You're going to see, in the future, they're not stuck at seven nanometer, obviously. They're good at manufacturing. They will continue to advance from seven and beyond. Now, is there 10x difference between five nanometer and seven nanometer? The answer is no. Architecture matters. Networking matters. That's why NVIDIA bought Melanox. Networking matters. Energy matters. And so all of that stuff matters. It's not simplistic like the way you're trying to steal it. We can move on from China. But it actually raises an interesting question about, we were discussing earlier these bottlenecks at TSMC and memory and so forth. And so if we're in this world where, you know, you're already the majority of N3, at some point, you'll be N2. You'll be the majority of that. Do you see that you could go back to N7, the spirit capacity of an older process node and say, hey, the demand for AI is so great and our capacity to expand the leading edge is not meeting it. So we're going to make a hopper or ampere about everything we know about Numerix today and all the other improvements you described. Do you see that world happening within before 2030? It's not necessary to. And the reason for that is because with every generation, the architecture, the architecture is more than just the transistor scale. It also, you're doing so much engineering and packaging and stacking and the Numerix and the system architecture. When you run out of capacity to easily go back to another node, that's a level of R&D that no one could afford. We could afford to lean forward. I don't think we could afford to go back. Now, if the world simply says, if on that day, on that day, let's do the thought experiment, on that day, we're just never going to have more capacity ever again. What I go back in U7 in a heartbeat. Of course, I would. One question somebody I was talking to had is, why Nvidia doesn't run multiple different chip projects at the same time with totally different architectures. You could do a cerebral style, wait for scale, you could do a dojo style, huge package, you could do one without CUDA. You have the resources and the engineering talent to do all of these in parallel. Why put all the eggs in one basket given Puno's where AI might go and architectures might go? Oh, we could. It's just that we don't have a better idea. Yeah, yeah, we could do all of those things. It's just not better. We simulate it all. They're in our simulator, provably worse. We wouldn't do it. We're working on exactly the projects that we want to work on. If the workload were to change dramatically, I don't mean the algorithms, I actually mean the work load. That depends on the shape of the market. We may decide to add other accelerators. Like for example, recently we added GROC and we're going to fold GROC into our CUDA ecosystem. We're doing that now because the value of tokens have gone up so high that you could have different pricing of tokens. Back in the old days, just a couple of years ago, tokens are either free or barely barely expensive. But now you can have different customers and those customers want different answers. And so because the customers make so much money, like for example our software engineers, if I can give them much more responsive tokens so that they're even more productive than they are today, I would pay for it. But that market is only recently emerged. And so I think that we now have we now have the ability to have the same model based on the response time have different segments. And that's the reason why we decided to expand the Pareto frontier and create a segment of inference that is faster response time even though it's lower, lower throughput. Until now, higher throughput is always better. We think that there could be a world where there could be very high ASP tokens. Even though the even though the throughput is lower in the factory, the ASP's make up for it. That's the reason why we did it. But otherwise, from an architecture perspective, I think in video architecture, I would rather put, if I have more money, I put more behind the architecture. I think this idea of extremely premium tokens and just the disaggregation of the inference market is very interesting. The segmentation. Yeah. Yeah. I find a question. Suppose a deep learning of revolution didn't happen. What would Nvidia be doing? Obviously, game is but given. Exorbit computing. Exorbit computing. The same thing we've been doing all along. The premise of our company is that Moore's Law is going to more general force computing. is good for a lot of things, but for a lot of computation is not ideal. So we combined an architecture called a GPU, CUDA, to a CPU so that we can accelerate the workload of the CPU. So different kernels of code or algorithms could be offloaded onto our GPU. As a result, you speed up an application by 100x, 200x. And where can you use that? Well, obviously engineering and science and physics and, you know, so on, so data processing, computer graphics, image generation, I mean, all kinds of things. Even if AI doesn't exist today, Nvidia will be very, very large. Yeah. And so, so I think the reason for that is fairly fundamental, which is the ability for general purpose computing to continue to scale, has largely run its course. And not the only way, but the way to do that is through domain specific acceleration. And one of the domain that we started with was computer graphics. But there are many, many other domains. I mean, there's, you know, all kinds of scientific particle physics and fluids and, you know, and so structured data processing, all kinds of different types of algorithms that benefit from CUDA. And so our mission was really to bring accelerated computing to the world. And advance the type of applications that general purpose computing can't do, and scale to the level of capability that helps break through certain fields of science. And so some of the early applications were molecular dynamics, seismic processing for energy discovery, and image processing, of course. And so all of those kinds of fields where, where general purpose computing is just simply too inefficient to do so. And so, yeah, if there's no AI, I would be very sad. But because of, because of, of the advances that we made in computing, we democratized deep learning. We made it possible for any researcher, or any researcher, or any scientist, anywhere, any student to be able to access a PC or, you know, a G4 is adding card and do amazing science. And that fundamental promise hasn't changed, not even a little bit. And so if you see GTS, if you watch GTS, there's the whole beginning part of it, none of its AI, that whole part of it with computational lithography or our quantum chemistry work, or, you know, all of that stuff, data processing work, all of that stuff is unrelated to AI. And it's still very important. I mean, I know that AI is very interesting and quite exciting. But, but there's a lot of people doing a lot of very important work that's not, not AI related, and tensors is not the only way that you compute with. And we want to help everybody. - Doesn't? - Thank you so much. - You're welcome. I enjoyed it. - Me too. - Speed.