Nvidia Part II: The Machine Learning Company (2006-2022)

[0] Still got Swedish House Mafia, Greyhound in my head from the pump up.

[1] Nice.

[2] It is funny how all, like, GPU companies, like I was watching a bunch of NVIDIA keynotes and AMD keynotes to get ready for this, and everyone is so, like, techno, neon lighting.

[3] Like, it's like crypto before crypto.

[4] Who got the truth?

[5] Is it you?

[6] Is it you?

[7] Is it you?

[8] Is it you?

[9] Is it you?

[10] Who got the truth now?

[11] Is it you?

[12] Is it you?

[13] Is it you?

[14] Welcome to Season 10, Episode 6 of Acquired, the podcast about great technology companies and the stories and playbooks behind them.

[15] I'm Ben Gilbert, and I am the co -founder and managing director of Seattle -based Pioneer Square Labs and our venture fund, PSL Ventures.

[16] And I'm David Rosenthal, and I am an angel investor based in San Francisco.

[17] And we are your hosts.

[18] When I was a kid, David, I used to stare into backyard bondfurt.

[19] fires and wonder if that fire flickering was doing so in a random way or if I knew about every input in the world, all the air, exactly the physical construction of the wood, all the variables in the environment, if it was actually predictable.

[20] And I don't think I knew the term at the time, but modelable.

[21] If I could know what the flame could look like if I knew all those inputs.

[22] And we now know, of course, it is indeed predictable, but the data and compute required to actually know that is extremely difficult.

[23] But that is what Nvidia is doing today.

[24] Ben, I love that intro.

[25] That's great.

[26] I was thinking like, where is Ben going with this?

[27] And this was occurring to me as I was watching Jensen sharing the omniverse vision for Nvidia and realizing Nvidia has really built all the building blocks, the hardware, the software for developers to use that hardware, all the user -facing software now and services to simulate everything in our physical world with their unbelievably efficient and powerful GPU architecture.

[28] And these building blocks, listeners, aren't just for gamers anymore.

[29] They are making it possible to recreate the real world in a digital twin to do things like predict airflow over a wing or simulate cell interaction to quickly discover new drugs without ever once touching a petri dish, or even model and predict how climate change will play out precisely.

[30] And there is so much to unpack here, especially in how NVIDIA went from making commodity graphics cards to now owning the whole stack in industries from gaming to enterprise data centers to scientific computing, and now even basically off -the -shelf, self -driving car architecture for manufacturers.

[31] And at the scale that they're operating at, these improvements that they're making are literally unfathomable to the human mind.

[32] And just to illustrate, if you are training one single speech recognition machine learning model these days.

[33] One, just one model.

[34] The number of math operations, like ads or multiplies, to accomplish it, is actually greater than the number of grains of sand on the earth.

[35] I know exactly what part of the research you got that from because I read the same thing and I was like, you got to be freaking kidding me. Isn't that nuts?

[36] I mean, there's just nothing better in all of the research that you and I both did.

[37] I don't think to better illustrate just the unbelievable scale of data and compute required to accomplish the stuff that they're accomplishing and how unfathomably small all of this is, the fact that that happens on one graphics card.

[38] Yep.

[39] So great.

[40] Okay, listeners, now is a great time to thank one of our big partners here at Acquired, ServiceNow.

[41] Yes, ServiceNow is the AI platform for business transformation, helping automate processes, improve service delivery, and increase efficiency.

[42] 85 % of the Fortune 500 runs on them, and they have quickly joined the Microsofts at the NVIDIAs as one of the most important enterprise technology vendors in the world.

[43] And, just like them, ServiceNow has AI baked in everywhere in their platform.

[44] They are also a major partner of both Microsoft and NVIDIA.

[45] I was at NVIDIA's GTC earlier this year, and Jensen brought up ServiceNow and their partnership many times throughout the keynote.

[46] So why is ServiceNow so important to both NVIDIA and Microsoft?

[47] companies we've explored deeply in the last year on the show.

[48] Well, AI in the real world is only as good as the bedrock platform it's built into.

[49] So whether you're looking for AI to supercharge developers and IT, empower and streamline customer service, or enable HR to deliver better employee experiences, service now is the platform that can make it possible.

[50] Interestingly, employees can not only get answers to their questions, but they're offered actions that they can take immediately.

[51] For example, smarter self -service for changing 401k contributions directly through AI -powered chat, or developers building apps faster with AI -powered code generation, or service agents that can use AI to notify you of a product that needs replacement before people even chat with you.

[52] With ServiceNow's platform, your business can put AI to work today.

[53] It's pretty incredible that ServiceNow built AI directly into their platform, so all the integration work to prepare for it that otherwise would have taken you years is already done.

[54] So if you want to learn more about the ServiceNow platform and how it can turbocharge the time to deploy AI for your business, go over to ServiceNow .com slash Acquired.

[55] And when you get in touch, just tell them Ben and David sent you.

[56] Thanks, ServiceNow.

[57] And after you finish this episode, come join the Slack, Acquired .fm slash Slack and talk about it with us.

[58] All right, David, without further ado, take us in.

[59] And as always, listeners, this is not investment advice.

[60] David and I may hold positions in securities discussed, and please do your own research.

[61] That's good.

[62] I was going to make sure that you said that this time because we're going to talk a lot about investing and investors in Nvidia stock over the years.

[63] It has been a wild, wild journey.

[64] So last we left our plucky heroes, Jensen Huang and Nvidia, in the end of our Nvidia, the GPU company years, ending kind of rough.

[65] flee, you know, 2004, 2005, 2006.

[66] They had cheated death, not once, but twice, the first time in the super overcrowded graphics card market when they were first getting started.

[67] And then once they sort of, you know, jumped out of that frying pan into the fire of Intel now gunning for them, coming to commoditize them like all the other, you know, PCI chips that plugged into the Intel motherboard back in the day.

[68] And they bravely, defend them off.

[69] They team up with Microsoft.

[70] They make the GPU programmable.

[71] This is amazing.

[72] They come out with programmable shaders with the G4s3.

[73] They power the Xbox.

[74] They create the CG programming language with Microsoft.

[75] And so here we are.

[76] It's now 2004, 2005.

[77] And this is a pretty impressive company.

[78] Public company stock is high flying after the tech bubble crash.

[79] They've conquered the graphics card market.

[80] Of course, there's ATI out there as well, which will come up again.

[81] But there's three pretty important things that I think the company built in the first 10 years.

[82] So one, we talked about this a lot last time, the six -month ship cycles for their chips.

[83] We talked about that, but we didn't actually say the rate at which they ship these things.

[84] I actually wrote down like a little list.

[85] So in the fall of 1999, they shipped the first G -Force card, the G -Forge -26.

[86] In the spring of 2000, G -Force 2, in the fall of 2000, G -4 -4 -2, in the fall of 2000, G -G -4 2 Ultra.

[87] Spring of 2001, G4s3, that's the big one with the programmable shaders.

[88] Then six months later, the G4s3 Ti -500.

[89] I mean, the normal cycle, I think we said, was two years, maybe 18 months for most of their competitors who just got largely left in the dust.

[90] Well, I was just thinking, you know, yeah, the competitors are gone at this point, but I'm thinking about Intel.

[91] How often did Intel ship new products, let alone fundamentally new architecture?

[92] You know, there was the 286 and then the 386 and the Pentio, and he got it to Pentio, I don't know, 5, whatever.

[93] Dude, I feel like the Intel product cycle is approximately the same as a new body style of cars.

[94] Yes, exactly.

[95] Every five, six years, there seems to be a meaningful new architecture change.

[96] And Intel is the driver of Moore's law, right?

[97] Like, these guys ship and bring out new architectures at warp speed.

[98] And they've continued that through to today.

[99] Two, one thing that we missed last time that is super important and becomes a big foundation of everything in video becomes today that we're going to talk about.

[100] They wrote their own drivers for their graphics cards.

[101] And we owed a big thank you for this and many other things to a great listener, very kind listener named Jeremy, who reached out to us in Slack and pointed us to a whole bunch of stuff, including the Asianometry YouTube channel.

[102] So good.

[103] I've probably watched like 25 Asianometry videos this week.

[104] So, so good.

[105] Huge.

[106] shout out to them.

[107] But all the other graphics cards companies at the time and most peripheral companies, they let the further downstream partners write the drivers for what they were doing.

[108] InVideo is the first one that said, no, no, no, we want to control this.

[109] We want to make sure consumers who use Nvidia cards have a good experience on whatever systems they're on.

[110] And that meant, A, that they could ensure quality.

[111] But B, they start to build up in the company, this base of really nitty -gritty, low -level software developers in this chip company.

[112] And there are not a lot of other chip companies that have capabilities like this.

[113] No. And what they're doing here is taking on a bigger fixed cost base.

[114] I mean, it's very expensive to employ all the people who are writing the drivers for all the different operating systems, all the different OEMs, all the different boards that has to be compatible with.

[115] But they viewed it as it's kind of an Apple -esque view of the world.

[116] we want the control or as much control as we can get over making sure that people using our products have a great user experience.

[117] So they were sort of willing to take the short -term pain of that expense for the long -term benefit of that improved user experience with their products.

[118] That their users, high -end gamers that want the best experience, you know, they're going to go out, they're going to spend the time $3, $500 on an NVIDIA top -of -the -line graphics card.

[119] They're going to drop it into the PC that they built.

[120] You know, they want it to work.

[121] I remember messing around with drivers back in the day and things not working.

[122] Like, this is super important.

[123] So all this is focusing.

[124] And then, of course, they have the third advantage in the company is programmable shaders, you know, which ATI copies as well.

[125] But like they innovated, like they've, you know, done all this.

[126] So all of this at this time, it's all in service of the gaming market.

[127] And one seed to plant here, David, when you say the programmable shaders, developers, the notion of a NVIDIA developer did not exist until this moment.

[128] it was people who wrote software that would run on the operating system and then from there maybe it would get that compute load would get offloaded to whatever the graphics card was but it wasn't like you were developing for the GPU for the graphics card with a language and a library that was specific to that card so for the very first time now they start to build a real direct relationship with developers so that they can actually start saying look if you develop for our specific hardware, there are advantages for you.

[129] And really, our specific gaming cards.

[130] Like everything we're talking about, these developers, they're game developers.

[131] All of this stuff, it's all in service to the gaming market.

[132] So, you know, again, they're a public company.

[133] They have this great deal with Microsoft.

[134] They bring out CG together.

[135] They're powering the Xbox.

[136] Wall Street loves them.

[137] They go from sub a billion dollar market cap company after the tech crash, up to five to six billion dollars, kind of by 2014.

[138] 2005, stock keeps going on a tear.

[139] By mid -2007, the stock reaches just under $20 billion market cap.

[140] You know, this is great.

[141] And this is all the stories.

[142] Like, this is pure play gaming.

[143] These guys have built such a great advantage in a developer ecosystem in a large and growing market, clearly, which is video games.

[144] Which on its own, that would be a great wave to surf.

[145] I mean, I think what's the gaming market today?

[146] 180 billion or something.

[147] And when we talked to Tripp Hawkins, who's sort like helped invent it or Nolan Bushnell, you know, it was zero then.

[148] And so Nvidia is sort of like on a wave that's at an amazing inflection point, they could totally just ride this gaming thing and be an important company.

[149] It's not running out of steam.

[150] I mean, like, how could you not be not just satisfied, but like more than satisfied with this as a founder?

[151] You're like, yes, I am the leading company in this major market, this huge wave that I don't see ending anytime soon.

[152] you know, 99 .9 % of founders who are themselves as a class, like, you know, very ambitious, are going to be satisfied with that.

[153] But not Jensen.

[154] But not Jensen.

[155] So while all this is happening, he starts thinking about, well, what's the next chapter?

[156] You know, I'm dominating this market.

[157] I want to keep growing.

[158] I don't want Envidio to be just a gaming company.

[159] So we ended last time with the little, you know, almost a surely apocryful story of a Stanford researcher.

[160] you know, sends the email to Jensen.

[161] And it's like, oh, you know, thanks to you, my son told me to go buy off the shelf, you know, G -Force cars at the local fries electronics, and I stuffed them into my PC at work.

[162] And, you know, I ran my models on this.

[163] He's a, I think it was a quantum chemistry researcher, supposedly.

[164] It was 10 times faster than the supercomputer I was using in the lab.

[165] And so thank you.

[166] I can get my life's work done in my lifetime.

[167] And Jensen loves that quote.

[168] It comes out at every GTC.

[169] So that story, if you're a skeptical listener, might beg two questions.

[170] First is a practical one.

[171] You know, we just said everything's about gaming here.

[172] And here's like a researcher, like a scientific researcher doing, you know, chemistry modeling using GForce cards for that.

[173] What's he writing this in?

[174] Well, it turns out.

[175] Progradable shaders, right?

[176] Yeah.

[177] They were shoehorning CG, which was built for graphics.

[178] They were translating everything.

[179] that they were doing into graphical terms, even if it was not a graphical problem they were trying to solve, and writing it in CG.

[180] This is not for the faint of heart, so to speak.

[181] Right.

[182] So everything is sort of metaphorical.

[183] He's a quantum chemistry researcher, and he's basically telling the hardware, okay, so imagine this data that I'm giving you is actually a triangle, and imagine that this way that I want to transform the data is actually like applying a little bit of lighting to the triangle, and then I want you to output something that you think is the right color pixel, and then I will translate it back into the result that I need for my quantum chemistry.

[184] You can see why that's suboptimal.

[185] Yeah.

[186] So he thinks this is an interesting market.

[187] He wants Nvidia to serve it.

[188] If you really want to do that right, it is a massive undertaking.

[189] It was 10 plus years to get to the company to this point.

[190] You know, what CG was is like a small sliver of the stack of what you would need to build for developers to use GPUs in a general purpose way like we're talking about.

[191] You know, it's kind of like they worked with Microsoft to make CG.

[192] It's like the difference between working on CG and like Microsoft building the whole dot net framework for developing on Windows or today, even better, Apple, right?

[193] Like everything Apple gives to iOS and Mac developers to develop on Mac.

[194] Right.

[195] Yeah, the analogy is not perfect, but it's like instead of just Apple saying, okay, Objective C is the way that you write code for our platforms, good luck.

[196] They're like, okay, well, will you need UI frameworks?

[197] So how about AppKit and Coco Touch?

[198] And how about all these other SDKs and frameworks like AR Kit and like StoreKit and like HomeKit?

[199] It's basically you need the whole sort of abstraction stack on top of the programming language to actually make it very accessible to write software for domains and disciplines that you know are going to be really popular using.

[200] that hardware.

[201] Exactly.

[202] So when Jensen commits himself and the company to pursuing this, he's biting off a lot.

[203] Now we talked about they've been writing their own drivers.

[204] So they have actually a lot of very low level, I don't mean low level like bad.

[205] I mean low level like infrastructure, like close, very difficult systems oriented programming talent within the company.

[206] So that kind of enables them to start here.

[207] But like still this is big.

[208] So then the second question, if you're a discerning investor, particularly in Nvidia, that you want to ask at this point in time is like, okay, Jensen, you're committing the company to a big undertaking.

[209] What's the business case for that?

[210] Show me the market.

[211] Don Valentine at this point would be sitting there listening to Jensen and being like, show me the market.

[212] And not only is it show me the market, but it's how long will the market take to get here?

[213] And it's how long is it going to take us and how many dollars and resources is it going to take us to actually get to something that's useful for that market when it materializes.

[214] Because while Kuda development began in 2006, that was not a useful, usable platform for six plus years at NVIDA?

[215] Yep.

[216] This is closer to on the order of the Microsoft development environment or the Apple development environment than what NVIDIA was doing before, which was like, hey, we made some APIs hasn't worked with Microsoft so that you can program for my thing.

[217] Right.

[218] I'm going to flash way forward just to illustrate the insane undertaking of this.

[219] I searched LinkedIn for people who work at Nvidia today and have the word CUDA in their title.

[220] There are 1 ,100 employees dedicated specifically to the CUDA platform.

[221] I'm surprised it's not 11 ,000.

[222] Yeah.

[223] Okay.

[224] So like where's the market for this?

[225] Yes, Ben, you asked the third question, which is, okay, the intersection of what does this take?

[226] to do this and when is the market going to get there in time and cost and all that.

[227] But even just put that aside, is there a market for this?

[228] Is the first order question?

[229] And the answer to that is probably no at this point in time.

[230] And what they're aiming at is scientific computing, right?

[231] It's researchers who are in science -specific domains who right now need supercomputers or access to a supercomputer to run some calculation that they think is going to take weeks or months, and wouldn't it be nice if they could do it cheaper or faster?

[232] Is that kind of the market they're looking at?

[233] Yeah, they're attacking like the cray market, like cray supercomputers, like that kind of stuff.

[234] You know, great company, right?

[235] But like, that's no Nvidia today.

[236] Right.

[237] And they were dominating the market.

[238] You know, yeah, it's scientific research computing.

[239] You know, it's drug discovery.

[240] It's probably a lot of this work they're thinking, oh, maybe we can get into more professional like Hollywood and architecture and other professional graphics domains.

[241] Yeah, yeah, yeah, sure.

[242] You know, you sum all that stuff up and, like, maybe you get to a couple billion dollar market, maybe, like, total market.

[243] And not enough to justify the time and the cost of what you're going to have to build out to go after this to any rational person.

[244] So, you know, here we come.

[245] Jensen and a video, like, they are doing this.

[246] He is committed.

[247] He's drunk the Kool -Aid.

[248] 2006, 2007, 2008.

[249] They are pouring a lot of resources into building what will become Kuda that will.

[250] we'll get to in a second.

[251] It already is good at this point in time.

[252] And I think Jensen's psychology here is sort of twofold.

[253] One is he is enamored with this market.

[254] He loves the idea that they can develop hardware to accelerate specific use cases in computing that he finds sort of fanciful.

[255] And he likes the idea of making it more possible to do more things for humanity with computers.

[256] But the other part of it is certainly a business model realization where he has spent the last gosh, at this point, 13, 14 years being commoditized in all these different ways.

[257] And I think he sees a path here to durable differentiation, where he's like, whoa.

[258] To own the platform.

[259] You know, it's kind of the Apple thing, again, to own the platform and to build hardware that's differentiated by not only software, but relationships with developers that use that custom software.

[260] Like, then I can build a really sort of like a company that can throw its weight around in the industry.

[261] 100%.

[262] Jensen, I don't know if he used it at the time because he probably would have gotten pilloryed, but maybe he did.

[263] I don't think he cared.

[264] He certainly has used it since.

[265] The way he thought about this was it, it wasn't just like, if we build it, they will come, which is what was going on.

[266] The phrase he uses is, if you don't build it, they can't come.

[267] So it's not even like, I'm pretty sure if we build it, they will come.

[268] It's one step removed from that.

[269] It's like, well, if we don't build it, they can't even possibly come.

[270] I don't know if they will come, but they can't come if we don't build it.

[271] So Wall Street is mostly willing to ignore this in 2006, 2007, 2008.

[272] The company's still growing really nicely.

[273] This great market cap run leading up to right before a financial crisis.

[274] But then, you know, you mentioned last time, I think it gets announced in 2006, maybe, and closes in 2007, AMD acquires ATI.

[275] And ATI was a very legit.

[276] competitors, the only standing legit competitor to Nvidia through its whole life.

[277] But now, AMD acquired it.

[278] And I think they acquired it for, what, $6, $7 billion, something like that?

[279] Something like that.

[280] So it was a lot of money.

[281] And then they put a lot of resources.

[282] Like, they weren't just acquiring this to, you know, get some talent.

[283] Like, they're like, no, no, this is going to be a big practical line for us.

[284] We're putting a lot of weight behind this.

[285] We haven't done the research into AMD the way we have into Nvidia.

[286] But the AMD radio online, which used to be the ATI Radio online, that is how you think about AMD as, and as a company is that they make these GPUs mostly for the gaming use case.

[287] Yep.

[288] Before the acquisition, I think the first PC I built in like end of high school, beginning of college, I think I had a radion card in it.

[289] I think I was probably in the minority.

[290] I think Nvidia was bigger, but for whatever reason I liked ATI at that point in time.

[291] So like, they were legit.

[292] Well, so here's Nvidia now focusing on this whole other thing.

[293] And you're still in the gaming market, which like we said is like, massive rising tide, your competitor now has all these resources and AMD that's fully dedicated to going after it.

[294] Mid -2008, Nvidia whiffs on earnings.

[295] Like, this is natural.

[296] They took their eye off the ball.

[297] Of course they did.

[298] And the stock gets just hammered.

[299] Because anything that Kuda empowers is not yet a revenue driver, and they've totally taken their eye off of gaming.

[300] Yes.

[301] So, you know, we said the high was around a $20 billion market cap.

[302] It drops 80 % 80.

[303] This isn't just the financial crisis.

[304] It's almost quaint, I think, you know, for me thinking back on the financial crisis now and like people freaking out the Dow, you know, the S &P dropping 5 % in a day and like, oh, that's a Thursday these days.

[305] It is literally the Thursday that we are recording.

[306] Yes.

[307] For a company stock to drop 80 % a technology company stock, even during the financial crisis, they're not just in the penalty box.

[308] They're like getting kicked to the curb.

[309] Right.

[310] Are they?

[311] done.

[312] The headlines at this point are is NVIDIA's run over.

[313] If you're most CEOs at this point in time, you're probably calling up Goldman or, you know, Allen Company or Frank Quattron and you're shopping this thing because how are you going to recover?

[314] But not Jensen.

[315] But not Jensen, obviously.

[316] So instead he goes and builds Kuda and continues to build Kuda.

[317] And this is, you know, just a sick context.

[318] Like, we get excited about a lot of stuff unacquired.

[319] But I think Kuda is like one of the greatest business stories of the last 10 years, 20 years more.

[320] I don't know.

[321] What do you think, Ben?

[322] I mean, I'd say it's one of the boldest bets we've ever covered, but so were programmable shaders.

[323] And so was NVIDIA's original attempt to make a more efficient quadrilateral focused graphics.

[324] Those were big bets.

[325] I think this is a bet on another scale, though.

[326] This is a bet that we don't cover that.

[327] off and on Acquire.

[328] Those were big bets relative to the company's size at the time, but this bet is like an iPhone -sized bet.

[329] That's exactly what this is.

[330] It's an iPhone -sized bet.

[331] It is a bet the company when you are already a several billion dollar company.

[332] Yes.

[333] An attempt to create something that if they are successful and this market materializes, this will be a generational company.

[334] Yep.

[335] So what is Kuda?

[336] It is Nvidia's compute, unified, device architecture.

[337] It is, as we've referred to, you know, thus far throughout the episode, a full, and I mean full development framework for doing any kind of computation that you would want on GPUs.

[338] Yeah, and in particular, it's interesting because I've heard Jensen reference it as a programming language.

[339] I've heard him reference it as a computing platform.

[340] It is all of these things.

[341] It's an API.

[342] It is an extension of C or C++.

[343] So there's a way that it's sort of a language, but importantly, it's got all these frameworks and libraries that live on top of it, and it enables super high -level application development, you know, really high abstraction layer development for hundreds of industries at this point to communicate down to Kudo, which communicates down to the GPU and everything else that they have done at this point.

[344] This is what's so brilliant.

[345] So right after we released, the same day that we were, the same day that we released part one.

[346] Yep.

[347] The first MVD episode we did a couple weeks ago, Ben Thompson had this amazing interview with Jensen on Stratory.

[348] And Jensen in this interview, I think, puts what Kuta is and how important it is, I think, better than I've seen anywhere else.

[349] So this is Jensen speaking to Ben.

[350] We've been advancing Kuda and the ecosystem for 15 years and counting.

[351] We optimize across the full stack, iterating between GPU, acceleration libraries, systems, and applications continuously, all while expanding the reach of our platform by adding new application domains that we accelerate.

[352] We start with amazing chips, but for each field of science, industry, and application, we create a full stack.

[353] We have over 150 SDKs that serve industries from gaming and design to life and earth sciences, quantum computing, AI, cybersecurity, 5G, and robotics.

[354] And then he talks about what it took to make this.

[355] This is like the point we we tried to like hammer home here.

[356] He says, you have to internalize that this is a brand new programming model, and everything that's associated with being a program processor company or a computing platform company had to be created.

[357] So we had to create a compiler team.

[358] We had to think about SDKs.

[359] We had to think about libraries.

[360] We had to reach out to developers and evangelize our architecture and help people realize the benefits of it.

[361] And we even had to help them market this vision so that there would be demand for their software that they write on our platform and on and on and on.

[362] It's crazy.

[363] It's amazing.

[364] And when he says that it's a whole new programming, I think he says maybe paradigm or way of programming.

[365] It is literally true because most programming languages up to this point and most computing platforms primarily contemplated serial execution of programs.

[366] And what Kuda did was it said, you know what, the way that our GPUs work and the way that they're going to work going forward is tons and tons of cores, all executing things at the same time, parallel programming, parallel architecture.

[367] Today, there's over 10 ,000 cores on their most recent consumer graphics card.

[368] So insanely, or dare I say, embarrassingly parallel, and Kuda is designed for parallel execution from the very beginning.

[369] That's the like catchphrase in the industry of embarrassingly parallel.

[370] And it's actually kind of a technical term.

[371] I don't know why it's embarrassing.

[372] It's basically the notion that this software is so parallelizable.

[373] which means that all of the computations that need to be run are independent.

[374] They don't depend on a previous result in order to start executing.

[375] It's sort of like it would be embarrassing for you to execute these instructions in order instead of finding a way to do it parallel.

[376] Ah, it's not that it's parallel that it's embarrassing.

[377] It's embarrassing if you were to do it the old way on CPUs, serially.

[378] I think that's the implication.

[379] Got it, got it.

[380] This is so obvious that it's embarrassingly parallel.

[381] Okay, now it makes sense.

[382] Now here's the coup de grasp.

[383] We're going to spend a few minutes talking about how brilliant this was.

[384] Everything we just described, this whole undertaking is like building the pyramids of Egypt or something here.

[385] It is entirely free.

[386] Invidia, to this day, now this may be changing.

[387] We'll talk about this at the end of the episode, has never charged a dollar for Kuda.

[388] But anyone can download it, learn it, use it, you know, blah, blah, blah.

[389] All of this work stand on the shoulders of everything in video has done.

[390] But, Ben, what is the but?

[391] It is closed source and proprietary exclusively to Nvidia's hardware.

[392] That's right.

[393] You do any of this work.

[394] You cannot deploy it on anything, but Nvidia chips.

[395] And that's not even just like, oh, Nvidia put in the like terms of service that you can't deploy this on, you know, AMD chips or whatever.

[396] Like literally doesn't work.

[397] Nope, it's full stack.

[398] It's like if you were to develop an iOS app and then try and deploy it on Windows, like, uh, It wouldn't work.

[399] It is integrated with the hardware.

[400] So OpenCL is sort of the main competitor at this point, and they do actually let OpenCL applications run on their chips, but nothing in Kuda is available to run elsewhere.

[401] It's so great.

[402] Okay, so now you can see this is just like Apple, and it's the Apple business model.

[403] Apple gives away all of this amazing platform ecosystem that they've built to developers, and then they make money by selling their hardware for very, very healthy gross margins.

[404] But this is why Jensen is so brilliant because back when they started down this journey in 2006, even before that when they started and then all through it, there was no iOS.

[405] There was no iPhone.

[406] Like it wasn't obvious that this was a great model.

[407] In fact, most people thought this was a dumb model that like Apple lost and the Mac was stupid a niche, and, like, Windows and Intel is what won, the open ecosystem.

[408] Well, but Windows and Intel did have proprietary development environments and, you know, full -stack dev tools.

[409] Oh, yeah.

[410] There's a lot of nuance here.

[411] It's not like they were, like, open source, per se.

[412] But it could run on any hardware.

[413] Well, except that it couldn't.

[414] It could only run on the Intel, IBM, Microsoft Alliance world.

[415] It wasn't running on PowerPCs.

[416] It wasn't running on anything Apple made.

[417] That's true.

[418] It's funny.

[419] In some ways, NVIDIA is like Apple.

[420] In other ways, they're like the Microsoft Intel IBM Alliance, except fully integrated with each other instead of being three separate companies.

[421] Yeah, that's maybe a good way to put it.

[422] It is sort of somewhere in between.

[423] There is nuance here.

[424] Remember when Clay Christensen was bashing on Apple in the early days of the iPhone being like, Open's going to win, Android's going to win, Apple is doomed, you know, clothes never works, you got to be modular.

[425] you can't be integrated.

[426] And like, you know, Clay was amazing and one of the greatest strategic, but I think that's just representative to me of like everybody thought that like the Apple model sucked.

[427] Yeah.

[428] I mean, it sucks unless you're at scale.

[429] And at the time, there was very little to believe that NVIDio was going to have the scale required to justify this investment or that there was a market to let them achieve the scale to justify this.

[430] That's the thing.

[431] Even if you were to say, okay, Jensen, I believe you, and I agree with you that this is a good model if you can pull it off.

[432] At the time, you could be Don Valentine or whoever looking around.

[433] And maybe Don was still looking around because they probably still held the stock.

[434] Being like, where's the market that's going to enable the scale you need to run this playbook?

[435] All right.

[436] So are you going to take us to 2011, 12?

[437] Where are we hopping back in here?

[438] If only the world were works like fiction and it were actually like a truly straight line.

[439] It's never a straight line.

[440] We will get there and that is what saves NVIDIA and makes this whole thing work.

[441] But they have some misadventures in between.

[442] So stock's getting hammered.

[443] It's 2008 and I'm just completely speculating on my own.

[444] But they're in the penalty box.

[445] they're committed to continuing to invest in CUDA and making general purpose computing on GPU a thing.

[446] I do wonder if they felt like, well, we got to do something to appease shareholders here.

[447] You know, we got a show that we're trying to be commercial here.

[448] So it's 2008.

[449] What's going on in 2008, you know, in the tech world?

[450] It's mobile.

[451] So in 2008, they launch the tegra chip.

[452] and platform with an NVIDIA.

[453] This may not be what saved the company.

[454] This is not what saved the company.

[455] This is more clown car style.

[456] Maybe that's too rough on NVIDIA.

[457] But what was Tegra?

[458] People might recognize that name.

[459] It was a full -on system on a chip for smartphones, competing directly with Qualcomm, with Samsung.

[460] Like, it was a processor, like an arm -based CPU, plus all of the other stuff you would need for a system on a chip to power.

[461] Android headsets, I mean, this is like a wild departure for it leverages none of Nvidia's core skill sets, except maybe graphics being part of smartphones, but like, come on, if there's ever a use case for integrated graphics, it's smartphones.

[462] Right, right, low power, smaller footprint.

[463] Yeah, totally.

[464] Do you know, this is one of my favorite parts about the whole research.

[465] Do you know what the first product was that shipped?

[466] using a tegra chip uh no it was the microsoft zoon hd media player uh that just tells you pretty much everything you need to know uh it did though the tegra system it is still around sort of to this day empowered the original tesla model s touchscreen so like before any of the autopilot, autonomous driving stuff, they were the processor powering just the infotainment, the touchscreen infotainment in the Model S. And I think that actually starts to help Nvidia get into the automotive market.

[467] The Tegra platform still to this day is the main processor of the Nintendo Switch.

[468] Oh, they repurposed it for that.

[469] Yeah, for that.

[470] And they, I think they still have their Nvidia Shield proprietary gaming device stuff that, I don't know that anybody buys those.

[471] Oh, this makes so much sense.

[472] because they basically have walked away from every console since the PlayStation 3.

[473] Yep.

[474] And so it's interesting that they have this thriving gaming division that doesn't power any of the consoles except the Nintendo Switch.

[475] And I always sort of wondered, like, why did they take on the Switch business?

[476] Because they kind of already had it done.

[477] It's not for the graphics cards.

[478] It was as somewhere to put the Tegras stuff.

[479] Fascinating.

[480] Quick aside, it's funny how these GPU companies have not been good at, transitioning to mobile.

[481] There's like a funny naming thing, but do you know what happened to...

[482] So there's the ATI Radion, which became the AMD Radion desktop series.

[483] They tried to make mobile GPUs.

[484] It didn't go great, and they ended up spinning that out and selling all that IP to another company.

[485] Do you know the company?

[486] Oh, I do not.

[487] Was it Apple?

[488] It is Qualcomm.

[489] and today is Qualcomm's mobile GPU division and Qualcomm's good at mobile and so it's a natural home for it.

[490] Do you know what that line of mobile GPU processors is called?

[491] No. It is the Ardino, A -R -D -E -N -O processors, and do you know why it's called the Ardino or Ardeno?

[492] No, that sounds super familiar, but no. The letters are rearranged from Radion.

[493] That's great.

[494] Yeah, that's great.

[495] So you're saying Nvidia's mobile graphics efforts didn't quite pan out.

[496] No. We didn't talk about this as much in the Sony episode, but my impression of the whole Android value chain ecosystem is that there's no profits to be made anywhere and Google keeps it that way on purpose.

[497] Ironically, they make a lot of money now on the Play Store.

[498] Ah, yeah, the Play Store, and ads.

[499] Right.

[500] I do think the primary way that they monetize it is not having to pay other people to acquire the search traffic.

[501] Right.

[502] But I mean for like partners, like if you are making everything from chips all the way up through hardware in the Android ecosystem, I don't think you're making it.

[503] Like maybe if you are the scale player, but like these things are designed to sell for dirt cheap as in products.

[504] Like there's no margin to be out of here.

[505] Yep.

[506] Yep.

[507] Also, before we continue, you just did the sidebar on the AMD mobile graphics chip.

[508] I see your sidebar.

[509] I'm going to raise you one more sidebar that we have to include that you know.

[510] because the NZS guys told us about this.

[511] So when NVIDA is going after mobile, they buy a mobile baseband company called I -SERA, British company called I -SERA in 2011.

[512] You know where I'm going with this.

[513] Oh, yes, I do.

[514] This is so good.

[515] It's a good seed plant to come back to later.

[516] You know, because they're investing in mobile integrity is going to be a thing, blah, blah, blah.

[517] And then a few years later, when they end up pretty much shutting down the whole thing, they shut down what they bought from IERA, they lay everyone off.

[518] The I -Sera found.

[519] who made a lot of money when NVIDIA bought them, they go off and they found a company called GraphCore that we're going to talk about a little bit at the end of the episode.

[520] It's maybe one of the primary sort of Nvidia Bear cases.

[521] Invidia bear cases, Nvidia killers out there.

[522] They've now raised about 700 million in venture capital and become mobile.

[523] In some ways, it's kind of like Bezos and Jet .com.

[524] Yes.

[525] if Jet had been successful.

[526] I think that's sort of the graph core to NVIDIA analogy.

[527] Yes.

[528] Well, I mean, a jury's still out if anybody's going to be really successful in competing with Nvidia, although I think the market now is probably, ironically, big enough that...

[529] Large.

[530] Yeah.

[531] Invidia can be the whale, and there can be plenty of big other companies, too.

[532] So, anyway, okay.

[533] Back to the story.

[534] So, NVIDIA is bumping along through all of this in the early late 2000s, early 2010s, you know, some years, growth is like 10 % maybe it's flat and others like this company is completely gone sideways in 2011 they whiff on earnings again stock goes through another 50 % drawdown it's cliche i was gonna say it i don't even know if you can say it about jensen like here we are the company is screwed again like everybody else would have given up but obviously not them so what happens basically a miracle happens.

[535] I don't know that there's any other way that you can describe this except like a miracle.

[536] So maybe this is actually not a great strategy case study of Jensen because it required a miracle.

[537] Well, Jensen would say it was intentional, that they did know the market timing and that the strategy was right and the investment was paying off and that they were doing this the whole time.

[538] Sure.

[539] In fact, even in the Ben Thompson interview, I think he said Ben basically lays out like, How did all these implausible things happen at exactly the right time?

[540] And his response is, oh, yes, we planned it all.

[541] It was so intentional.

[542] Jensen did not plan AlexNet or see it coming because nobody saw AlexNet coming.

[543] So in 2009, a Princeton computer science professor and also undergrad alum of Princeton, just like yours truly.

[544] Woo, a wonderful place named Fei -Fei Lee.

[545] Their specialty is artificial intelligence and computer science, starts working on an image classifying project.

[546] that she calls ImageNet.

[547] Now, the inspiration for this was actually a way old project from, I think, the 80s at Princeton called WordNet.

[548] That was like classifying words.

[549] This is classifying image.

[550] ImageNet.

[551] And her idea is to create a database of millions of labeled images, like images that they have a correct label applied to them, like this is a dog or this is a strawberry or something like that.

[552] and that with that database, then artificial intelligence image recognition algorithms could run against that database and see how they do.

[553] So like, oh, look at this image of, you know, you and I would look at it.

[554] I'd be like, that's a strawberry.

[555] But you don't give the answer to the algorithm and the algorithm figures out if it thinks it's a strawberry or a dog or whatever.

[556] So she and our collaborators start working on this.

[557] It's super cool.

[558] They build the database.

[559] They use a mechanical Turk, Amazon Mechanical Turk, to build it.

[560] And then, And one of them, I'm not exactly sure who, if it was Faye Faye or somebody else, has the idea of like, well, you know, we've got this database.

[561] We want people to use it.

[562] Well, let's make a competition.

[563] This is like a very standard thing in computer science academia of like, let's have a competition, an algorithm competition.

[564] So we'll do this annually.

[565] And anyone, any team can submit their algorithms against the ImageNet database.

[566] And they'll compete.

[567] Like who can get the lowest error rate, like the most number of images.

[568] Percentage of the image is correct.

[569] And this is great.

[570] So it brings her great renown, becomes popular in the AI research community.

[571] She gets poached away by Stanford the next year.

[572] I guess that's okay because I went there too, so that's fine.

[573] And she's still there to the...

[574] I know.

[575] I couldn't resist.

[576] I couldn't resist.

[577] I just...

[578] She's like a kindred spirit to me. Do you know?

[579] I know you do know, but I bet most listeners do not know what her endowed tenure chair is at Stanford today.

[580] I do.

[581] She is the Sequoia chair.

[582] Yes, the Sequoia Capital Professor of Computer Science at Stanford.

[583] So cool.

[584] Why does she become the Sequoia capital chair?

[585] And what does all this have to do with Invidia?

[586] Well, in the 2012 competition, a team from the University of Toronto submits an algorithm that wins the competition.

[587] And it doesn't just win it by like a little bit.

[588] It wins it by a lot.

[589] So the way they measure this is 100 % of the images in the database.

[590] What percentage of them did you get wrong?

[591] So it wins it by over 10%.

[592] I think it had a 15 % error rate or something in the next.

[593] Like all the best previous ones have been like 25.

[594] something percent.

[595] Yes.

[596] This is like someone breaking the four minute mile.

[597] Actually, in some ways it's more impressive than the four minute mile thing because they just didn't brute force their way all the way there.

[598] They like try to completely different approach.

[599] Yes.

[600] And then boom showed that we could get way more accurate than anyone else ever thought.

[601] So what was that approach?

[602] Well, they called the team, which was composed of Alex Krizepske, was the primary lead of the team.

[603] He was a PhD student in collaboration with Ilya Sutskiver and Jeff Hinton.

[604] Jeff Hinton was the PhD advisor of Alex.

[605] They call it AlexNet.

[606] What is it?

[607] It is a convolutional neural network, which is a branch of our artificial intelligence called deep learning.

[608] Now deep learning is new for this use case, but Ben is you weren't exactly right.

[609] It had been around for a long time, a very long time.

[610] And deep learning neural networks, this was not a new idea.

[611] The algorithms had existed for many decades, I think, but they were really, really, really computationally intensive.

[612] They required to train the models to do a deep neural network.

[613] You need a lot of compute, like on the order of, you know, like grains of sand that exist on Earth.

[614] It was completely impossible with a traditional computer architecture that you could make these work in any practical applications.

[615] And people were forecasting, too, like, when with Moore's Law?

[616] When will we be able to do this?

[617] And it still seemed like the far future, because not only did Moore's Law need to happen, but you also needed the NVIDIA approach of massively parallelizable architecture, where suddenly you could get all these incredible performance gains, not just because you're putting, you know, more transistors in a given space, but because you're able to run programs in parallel now.

[618] Yes.

[619] So Alex Nat took these old ideas and it implemented them on GPUs.

[620] And to be very specific, it implemented them in Kuda on Nvidia GPUs.

[621] We cannot overstay the importance of this moment, not just for Nvidia, but for like computer science, for technology, for business, for the world, for us staring at the screens of our phones all day every day.

[622] This was the big bang moment for artificial intelligence and Nvidia and Kuda were right there.

[623] Yep.

[624] It's funny.

[625] There's another example within the next couple of years, 2012, 2013, Nvidia had been thinking about this notion of general purpose computing for their architecture for a long time.

[626] In fact, they even thought about should we relaunch our GPUs as GPGPUs, general purpose graphics processing units.

[627] And of course, they decided not to do that, but just built Kuda.

[628] Which is code word for like, we've been searching for years for a market for this thing.

[629] We can't find a market.

[630] So we'll just say, you can do it for anything.

[631] Right.

[632] And so deep learning is generating a lot of buzz, you know, a lot from this AlexNet competition.

[633] And so in 2013, Brian Katel Zanzaro, who's a research scientist at NVIDIA, published a paper with some other researchers at Stanford, which included Andrew Ng, where they were able to take this unsupervised learning approach that had been done inside the Google Brain Team, where they had sort of, the Google brain team had sort of published their work on this, and it had a thousand nodes, and you know, this is a big part of the sort of early neural network hype cycle of people trying cool stuff, and this team was able to do it with just three nodes.

[634] So totally different model, super parallelized, lots of compute for a super short period of time in a really high performance computing way, or HPC, as it would sort of become known.

[635] And this ends up being the very core of what becomes QDN, which is the library for deep neural networks that's actually baked into CUDA that makes it easy for data scientists and research scientists everywhere who aren't hardware engineers or software engineers to just pretty easily write high -performance deep neural networks on Nvidia hardware.

[636] So this AlexNet thing, plus then Brian and Andrew Ing's paper, it just collapses all these sort of previously thought to be impossible lines to cross and just makes it way easier and way more performant and way less energy intensive for other teams to do it in the future.

[637] Yep.

[638] And specifically to do deep learning.

[639] So I think at this point, like everybody knows that this is pretty important, but it's not, that much of a leap to say if you can train a computer to recognize images on its own, that you can then train a computer to see on its own, to drive a car on its own, to play chess, to play go, to make your photos look really awesome when you take them on the latest iPhone, even if you don't have everything right.

[640] To eventually let you describe a scene and then have a transformer model paint that scene for you in a way that is unbelievable that a human didn't make it.

[641] Yep.

[642] And then most importantly, for the market that Jensen and NVIDIA are looking for, you can use the same branch of AI to predict what type of content you might like to see next show up in your feed of content and what type of ad might work really, really, really well on you.

[643] So basically all of these people we were just talking about I bet a lot of you recognize their names.

[644] They get scooped up by Google.

[645] Fay -Fei Lee goes to Google.

[646] Brian went to Baidu, and he's back at Nvidia now doing Applied AI.

[647] Brian went to Baidu.

[648] Jeff Hinton goes to Facebook.

[649] So, you know, all the other markets, like even throw out, say you don't believe in self -driving cars, you don't think it's going to happen or any of this other stuff.

[650] Like, it doesn't matter.

[651] Like, the market of advertising, of digital advertising that this enables is a freaking multi -trillion dollar market.

[652] And it's funny because, like, that feels like, ooh, that's the killer use case.

[653] But that's just the easiest use case.

[654] That's the most, like, obvious, well -labeled data set that these models don't have to be amazingly good because they're not generating unique output.

[655] They're just assisting in making something more efficient.

[656] But then, like, flash forward 10 more years, and now we're in these crazy transformer models with, I don't know if it's hundreds of millions or billions of parameters, things that we thought only humans could do are now being done by machines and it's like it's happening faster than ever.

[657] Yep.

[658] So I think to your point, David, it's like, oh, there was this big cash cow enabled by neural networks and deep learning in advertising.

[659] Sure, but that was just the easy stuff.

[660] Right, but that was necessary, though.

[661] This was finally the market that enabled the building of scale and the building of technology to do this.

[662] And in the Ben Thompson, Jensen interview, Ben, actually says this when he's sort of realizing this talking to Jensen, he says, this is Ben talking, the way value accrues on the internet in a world of zero marginal costs where there's just an explosion and abundance of content, that value accrues to those who help you navigate the content.

[663] He's talking about aggregation theory, duh.

[664] And then he says, what I'm hearing from you, Jensen, is that yes, the value accrues to people that help you navigate that content.

[665] But someone has to make the chips and the software so that they can do that effectively.

[666] And it's like it sort of used to be with Windows was the consumer facing layer and Intel was the other piece of the Wintel monopoly.

[667] This is Google and Facebook and a whole list of other companies on the consumer side and they're all dependent on NVIDIA.

[668] And that sounds like a pretty good place to be.

[669] And indeed it was a pretty good place to be.

[670] Amazing place to be.

[671] Oh my gosh.

[672] The thing is like the market did not realize this for years.

[673] And I mean, I didn't realize this.

[674] And you probably didn't realize this.

[675] We were the class of people working in tech as venture capitalists that should have.

[676] Ooh, do you know the Mark Andreessen quote?

[677] Oh, no. Oh, this is awesome.

[678] Okay, so it's a couple years later.

[679] So it's like getting more obvious, but it's 2016.

[680] And Mark Andreessen gave an interview.

[681] He said, we've been investing in a lot of companies applying deep learning to many areas.

[682] And every single one effectively comes in building on Nvidia's platforms.

[683] It's like when people were all building on Windows in the 90s or all building on the iPhone in the late 2000s.

[684] And then he says, for fun, our firm has an internal.

[685] game of what public companies we'd invest in if we were a hedge fund.

[686] We'd put in all of our money to Nvidia.

[687] This is like, it was paradigm, right, that called all of their capital in one of their funds and put it into Bitcoin when it was like $3 ,000 a coin or something like that.

[688] We all should have been doing this.

[689] So literally, Nvidia stock in 20, like recent, like this is now known, 2012, 13, 14, 15.

[690] It doesn't trade above like five bucks a share.

[691] And Nvidia, today, as we record this is, I think, about 220 a share.

[692] The high in the past year has been well over 300.

[693] Like, if you realized what was going on, and again, in a lot of those years, it was not that hard to realize what was going on.

[694] Wow.

[695] Like, it was huge.

[696] It's funny.

[697] So there was even, and we'll get to what happened in 2017 and 2018 with crypto in a little bit, but there was a massive stock run up to like $65 a share in 2018.

[698] And even as late as I think the very beginning of 2019, you could have gotten it.

[699] I tweeted this, and we'll put the graph on the screen and the YouTube version here, you could have gotten it in that crash for 34 bucks a share.

[700] In 2019.

[701] If you zoom out on that graph, which is the next tweet here, that you can see that like, in retrospect, that little crash just looks like nothing.

[702] You don't even pay attention to it in the crazy run -up that they had to 350 or whatever their all -time high was.

[703] Yeah, it's wild.

[704] A few more wild things about this.

[705] It's not an 2016.

[706] Again, AlexNet happens in 2012.

[707] It's not until 2016 that Nvidia gets back to the $20 billion market cap peak that they were in 2007 when they were just a gaming company.

[708] That's almost 10 years.

[709] I really hadn't thought about it the way that you're describing it.

[710] But the breakthrough happened in 2010, 2011, 2012.

[711] Lots of people had the opportunity, especially because freaking Jensen's talking about it on stage.

[712] He's talking about our earnings calls at this point.

[713] He's not keeping this a secret.

[714] No, he's like trying to tell us all that this is the future.

[715] And people are still skeptical.

[716] Everyone's not rushing to buy the stock.

[717] We're watching this freaking magic happen using their hardware, using their software on top of it.

[718] And like even semiconductor analysts who are like students of listening to Jensen talk and following the space very closely, sort of think he sounds like a crazy person when he's up there espousing that the future is neural networks and we're going to go all in.

[719] and we're not pivoting the business, but from the amount of attention that he's giving in earnings calls to this versus gaming, I mean, everyone's just like, are you off your rocker?

[720] I think people had just lost trust and interest, you know, after like, there were so many years of like, they were so early with Kuda and early taking out again.

[721] They didn't even know that this, like they didn't know AlexNet was going to happen.

[722] Right.

[723] Jensen felt like the GPU platform could enable things that the CPU paradigm could not and he had this faith that something would happen but you didn't know this was going to happen and so for years he was just saying that like we're building it they will come you know and to be more specific it was that well look the GPU has accelerated the graphics workload so we've taken the graphics workload off of the CPU the CPU is great it's your primary workforce for all sorts of flexible stuff but we know graphics needs to happen in its own separate environment and have all these fancy fans on it and get super cooled.

[724] And it needs these matrix transforms.

[725] The math that needs to be done is matrix multiplication.

[726] And there was starting to be this belief that like, oh, well, because the, you know, professor, the apocryphal professor told me that he was able to use these program.

[727] The matrix transforms to work for him.

[728] You know, maybe this matrix math is really useful for other stuff.

[729] And sure, it was for scientific computing.

[730] And then, honestly, like, it fell so hard into NVIDIA's lap, that the thing that made deep learning work was massively parallelized matrix math.

[731] And they're like, NVIDIA is just like stirring down at their GPUs, like, I think we have exactly what you are looking for.

[732] Yes.

[733] There's that same interview with Brian Katazaro, he says about when all this happened.

[734] He says, the deep learning happened to be the most important of all applications that need high throughput computation, understatement of the century.

[735] And so once Nvidia saw that, it was basically instant.

[736] The whole company just latched onto it.

[737] There's so many things to Law Jensen for, you know, he was painting a vision for the future, but he was paying very close attention.

[738] And the company was paying very close attention to anything that was happening.

[739] And then when they saw that this was happening, they were not asleep at the switch.

[740] Yeah, 100%.

[741] It's interesting thinking about the fact.

[742] fact that in some ways it feels like an accident of history, in some ways it feels so intentional that graphics is an embarrassingly parallel problem because every pixel on a screen is unique.

[743] I mean, they don't have a core to drive every pixel on the screen.

[744] There's only 10 ,000 cores on the most recent Nvidia graphics cards, but there's not, which is crazy, right, but there's way more pixels on a screen.

[745] So, you know, they're not all doing every single pixel at the same time, every clock iteration.

[746] But it worked out so well that neural networks also can be done entirely in parallel like that, where every single computation that is done is independent of all the other computations that need to be done.

[747] So they also can be done on this super parallel set of cores.

[748] It's just, you got to wonder, like, when you kind of reduce all this stuff to just math, It is interesting that these are two very large applications of the same type of math.

[749] In the search space of the world of what other problems can we solve with parallel matrix multiplication, there may be more.

[750] There may even be bigger markets out there.

[751] Totally.

[752] Well, I think they probably will be a big part of Jensen's vision that he paints for Nvidia now, which we'll get to in a sec, is this is just the beginning.

[753] You know, there's robotics, there's autonomous vehicles, there's the omniverse.

[754] It's all coming.

[755] It's funny.

[756] We just joked about how like nobody saw this before the run up in 2016, 2017.

[757] There were all these years where like Mark Andreessen knew, you know, whether he made money in his personal account or not, you know, we'll have to ask him.

[758] But then in 2018, another class of problems that are embarrassingly paralyzable is, of course, cryptocurrency mining.

[759] And so a lot of people were going out and buying consumer Nvidia, you know, graphics cards and using them to set up crypto mining rigs in 2016 and 2017.

[760] And then when the crypto winter hit in 2018 and the end of the ICO craze and all that, the mining rig demand fell off.

[761] And this has become so big for Nvidia that their revenue actually declined.

[762] Right.

[763] Yeah.

[764] So a couple interesting things here.

[765] Let's talk about technically why.

[766] So the way crypto mining works is effectively guess and check.

[767] You're effectively brute forcing an encryption scheme.

[768] And when you're mining, you know, you're trying to discover the answer to something that is hard to discover.

[769] So you're guessing, if that's not the right thing, you're incrementing, you're guessing again.

[770] And that's a vast oversimplification and not technically exactly right, but that's the right way to think about it.

[771] And if you were going to guess and check at a math problem, and you had to do that on the order of a few million times in order to discover the right answer, you could very unlikely discover the right answer on the first time, but, you know, that probabilistically is only going to happen to you once if ever.

[772] And so, well, the cool thing about these chips is that, A, they have a crap ton of cores.

[773] So the problem like this is massively parallelizable because instead of guessing and checking with one thing, you can guess and check with 10 ,000 at the same time, and then 10 ,000 more, and then 10 ,000 more.

[774] And the other thing is, it is matrix math.

[775] So yet again, there's this third application beyond gaming, beyond neural networks, there's now this third application in the same decade for the two things that these chips are uniquely good at.

[776] And so it's interesting that, like, you could build hardware that's better for crypto mining or better for AI, and both of those things have been built by Nvidia and their competitors now, but the sort of like general purpose GPU happen to be pretty darn good at both of those things.

[777] Well, at least way, way, way better than a CPU.

[778] Yeah.

[779] As some of Nvidia's startup competitors put it today, and Cerebris is the one that I'm thinking of, they sort of say, well, the GPU is a thousand times better, or, you know, much, much better than a CPU for doing this kind of stuff.

[780] But it's like a thousand times worse than it should be.

[781] There exists much more optimal solutions for, you know, doing some of this AI stuff.

[782] Interesting.

[783] Really begs the question of, like, how good is good enough in these use cases.

[784] Right.

[785] And now, I mean, to flash way forward, the game that Nvidia and everyone else all these upstarts are playing is really, it's still the accelerated computing game, but now it's how do you accelerate workloads off the GPU instead of off the CPU?

[786] Interesting.

[787] Well, back to Crypto Winter.

[788] The Nvidia stock gets hammered again.

[789] It goes through another 50 % drawdown.

[790] This is just like every five years, this has got to happen.

[791] Which is fascinating because at the end of the day, it was a thing completely outside their control.

[792] Like, people were buying these chips for for a use case that they didn't build the chips for.

[793] They had really no idea what people were buying them for.

[794] So it's not like they could even get really good market channel intelligence on are we selling to crypto miners or are we selling to, you know, people that are going to use these for gaming.

[795] Right.

[796] They're selling the Best Buy and then people go buy them and Best Buy.

[797] Right.

[798] And some people are buying them wholesale, like if you're actually starting a data center to mine.

[799] But a lot of people are just doing this in their basement with consumer hardware.

[800] So they don't have perfect information on this.

[801] And then, of course, the price crashing makes it either unprofitable or less profitable to be a minor.

[802] And so then your demand dries up for this thing that you, A, didn't ask for, and B, had poor visibility into knowing if people were buying in the first place.

[803] So the management team just looks terrible to the street at this point, because they had just no ability to understand what was going on in their business.

[804] And I think a lot of street still was still a this hangover of skepticism about this deep learning thing like what jensen okay and so it was kind of any excuse to sell off it took but anyway that was short live to the 50 % dip because with the use case and specifically the enterprise use case for GPUs for deep learning like it just takes off and so this is really interesting if you look at invidia's um they report financials a couple different ways, but one of the ways they break it out is a few different segments is the gaming consumer segment and then their data center segment.

[805] And it's like data center.

[806] Like what are they in the data set?

[807] Well, all the uses for, right, all of the stuff we're talking about, it's all done in the data center.

[808] Like, Google isn't going and buying, you know, a bunch of Nvidia GPUs and hooking them up to the laptops of their software engineers.

[809] Like, is Stadia still a thing?

[810] Like, I think that's used for cloud gaming and some, like there are.

[811] But if it's all happening in the data center is my point.

[812] Right, right.

[813] I guess what I'm saying, my argument is every time I see data center revenue, in my mind, I sort of make it synonymous with this is their ML segment.

[814] Ah, yes, yes, that's what I'm saying.

[815] I agree.

[816] Yeah.

[817] Now, the data center, this is really interesting, again, because they used to sell these cards that would get packaged, put on a shelf, a consumer would buy them.

[818] Yeah, they made some specialty cards for the scientific computing market and stuff like that.

[819] But this data center opportunity, like, man, do you know the prices that you can sell gear to data centers for?

[820] Like, it makes the RTX 3090 look like a pittance.

[821] And the RTX 3090, which is their most expensive high -end graphics card that you can buy as a consumer, it was $3 ,000.

[822] Now it's like $2 ,000.

[823] But if you're buying, I don't know, what's the latest?

[824] It's not the A100.

[825] It's the H -100.

[826] So the A -100, they just announced the H -100.

[827] And that's what, like 20 or 30 grand in order to just get one card?

[828] Yeah.

[829] And people are buying a lot of these things.

[830] Yeah, it's crazy.

[831] It's crazy.

[832] It's funny.

[833] I tweeted about this.

[834] And I was sort of wrong, but then like everything, there's nuance.

[835] You know, Tesla has announced making their own hardware.

[836] They're certainly doing it for the on the car, the inference stuff, like the full self -driving computer on Teslas.

[837] They now make those chips themselves.

[838] The Tesla Dojo, which is the training center, that they announced they were also going to make their own silicon for that they actually haven't done it yet so they're still using invidia chips for their training the current compute cluster that they have that they're still using i want to say i did the math and like assumed some pricing i think they spent between 50 and 100 million dollars that they paid invidia for all of the compute in that well that's one customer that's one customer that's one customer for one use case at that one customer crazy I mean, you see this show up in their earnings.

[839] So we're at the part of the episode where we're close enough to today that it's best illustrated by the today numbers.

[840] So I'll just flash forward to what the data center segment looks like now.

[841] So two years ago, they had about $3 billion of revenue, and it was only about half of their gaming revenue segment.

[842] So gaming, you know, through all this, through 2006 to AlexNet, all the way, you know, another decade forward to 2020, gaming is still king.

[843] It generates almost $6 billion in revenue.

[844] the data center revenue segment was $3 billion, but had been pretty flat for a couple of years.

[845] So then insanely, over the last two years, it 3xed, the data center segment 3xed.

[846] It is now doing over $10 .5 billion a year in revenue, and it's basically the same size as the gaming segment.

[847] It's nuts.

[848] It's amazing how it was like sort of obvious in the mid -2010s, but when the enterprises really showed up and said, we're buying all this hardware.

[849] and inputting it in our data centers.

[850] And whether that's the hyperscalers, the cloud folks, Google, Microsoft, Amazon, putting it in their data centers, or whether it's companies doing it in their own private clouds or whatever they want to call it these days, on -prem data centers, everyone is now using machine learning hardware in the data center.

[851] Yep.

[852] And Nvidia is selling it for very, very, very healthy gross margins, Apple level gross margins.

[853] Yes, exactly.

[854] So speaking of the data center, a couple things.

[855] One, in, this is so, InVIDIA, in 2018, they actually do change the terms of the user agreements of their consumer cards of G -Force cards that you cannot put them in data centers anymore.

[856] They're like, we really do need to start segmenting a little bit here.

[857] And we know that the enterprises have much more willingness to pay.

[858] And it is worth it.

[859] I mean, you buy these crazy data center cards and they have like twice as many transistors.

[860] actually, they don't even have video outputs.

[861] Like, you can't use the Data Center GPUs, like the A100 does not have video out.

[862] So they actually can't be used as graphic cards.

[863] Oh, yeah.

[864] There's a cool Linus Tech Tips video about this, where they get a hold of an A100 somehow, and then they run some benchmarks on it, but they can't actually, like, drive a game on it.

[865] Oh, fascinating.

[866] Yeah, so fun.

[867] Data Center stuff is, like, super high horsepower, but of course, like, useless to run a game on because you can't pipe it to a TV or a monitor.

[868] But then it's interesting that they're sort of artificially doing it the other way around and saying, for those of you who don't want to spend $30 ,000 on this and are trying to like make your own little rig at home, your own little data center rig at home, no, you cannot rack these things.

[869] Don't think about going to fries and buying a bunch of two forces.

[870] Ironic because that's how the whole thing started.

[871] But anyway, in 2020, they acquire an Israeli data center compute company called Melanox that I believe focuses on a networking compute within the data center for about $7 billion, integrate that into their ambitions in building out the data center.

[872] And the way to think about what Melanox enables them to do is now they're able to have super high bandwidth, super low latency connectivity in the data center between their hardware.

[873] So at this point, they've got NVLink, which is their, it's like the, what does Apple call it, a proprietary interconnect, or I think AMD calls it the infinity fabric.

[874] It's the super high bandwidth chip -to -chip connection.

[875] So think about what Melanox lets them do is it lets them have these extremely high bandwidth switches in the data center to then let all of these different boxes with Nvidia hardware and them communicate super fast to each other.

[876] That's awesome because, of course, these data centers, that's the other thing about, you know, customers like that Tesla example I gave.

[877] They're not buying cards, the enterprise customers.

[878] They're buying, solutions from NVIDIA.

[879] They're buying big boxes with lots of stuff in them.

[880] You say solutions?

[881] I hear gross margin.

[882] That's such a great quote.

[883] We should like frame that and put it on the wall of the acquired museum.

[884] It is true.

[885] Acquiring Melanox not only like enables this now we have the super high connectivity thing, but this is what leads to this introduction of this third leg of the stool of computing for NVIDIA that they talk about now, which is you had your CPU.

[886] It's great.

[887] It's your workhorse.

[888] You know, it's your general purpose computer.

[889] Then there's the GPU, which is really a GPGPU that they've really beefed up.

[890] And they've really like, for the enterprise, for these data centers, they've put tensor cores in it to do the machine learning specific 4x 4x multiplication super fast and do that really well.

[891] And they've put all this other non -gaming data center specific AI modules onto these chips and this hardware.

[892] And now what they're saying is, you've got your CPU, you've got your GPU.

[893] Now there's a DPU.

[894] And this data processing unit that's like kind of borne out of the Melanox stuff is the way that you really efficiently communicate and transform data within data centers.

[895] So the unit of how you think about, like the black box just went from a box on a rack to now you can kind of think about your data center as the black box.

[896] And you can write at a really high abstraction layer, and then Nvidia will help handle how things move around the data center.

[897] All right, listeners, our sponsor is one of our favorite companies, Vanta, and we have something very new from them to share.

[898] Of course, you know Vanta enables companies to generate more revenue by getting their compliance certifications.

[899] That's SOC2, ISO -27 -01.

[900] But the thing that we want to share now is Vanta has grown to become the best security compliance platform as you hit hypergrowth and scale into a larger enterprise.

[901] It's kind of wild.

[902] When we first started working with Vanta and met Christina, my gosh, they had like a couple hundred customers, maybe.

[903] Now they've got 5 ,000, some of the largest companies out there.

[904] It's awesome.

[905] Yeah.

[906] And they offer a tremendous amount of customization now for more complex security needs.

[907] So if you're a larger company and in the past, you showed Vanta to your compliance department, you might have heard something like, oh, well, we've already got a compliance process in place and we can't integrate this new thing.

[908] But now, even if you already have a SOC2, Vanta makes maintaining your compliance even more efficient and robust.

[909] They launched vendor risk management.

[910] This allows your company to quickly understand the security posture of the vendors that you're choosing in a standardized way that cuts down on security review times.

[911] This is great.

[912] And then on the customization front, they now also enable custom frameworks built around your controls and policies.

[913] Of course, that's in addition to the fact that with Vanta, you don't just become compliant once, you stay cool.

[914] compliant with real -time data pulled from all of your systems, now all of your partner's systems, and you get a trust report page to prove it to your customers.

[915] If you click the link in the show notes here or go to vanta .com slash acquired, you can get a free trial.

[916] And if you decide you love it, you will also get $1 ,000 off when you become a paying customer.

[917] Make sure you go to vanta .com slash acquired.

[918] Okay, so I said one more thing on the data center.

[919] Yes.

[920] That one more thing is, it's easy to forget now.

[921] I know because we've just been deep on this.

[922] Invita was going to buy Arm.

[923] Do you remember this?

[924] Yes, they were.

[925] And in fact, this is going to be like a corporate communications nightmare.

[926] Everyone out there, Jensen, their IR person, different tech people who are being interviewed on various podcasts, were talking about the whole strategy and how excited they are to own Arm and how Nvidia is going to be, you know, it's good on its own, but it could be so much better if we had Arm and here's all the cool stuff we're going to do with it.

[927] And then it doesn't happen.

[928] They were talking about it like it was a done deal.

[929] And now you've got dozens of hours of people talking about the strategy.

[930] So you're almost like, it's funny that now, after listening to all that, I'm sort of like disappointed with NVIDIA's ambition on its own without having the strategic assets of ARM.

[931] Yeah.

[932] We should revisit Arm at some point.

[933] We did do the soft bank acquiring ARM episode years and years ago now.

[934] But, you know, you think Arm, like they are a CPU architecture company whose primary use cases.

[935] is mobile and smartphones.

[936] It's like everything that Intel screwed up on back in the misguided mobile era.

[937] Now they're going and buying like the most important company in that space.

[938] You know, and it's interesting.

[939] Like again, in the Ben Thompson interview, Jensen talks all about this.

[940] And maybe this is just justifying in retrospect, but I don't think so.

[941] He's like, look, it was about the data center.

[942] Yeah, like everything Arm does is like great and that's fine.

[943] But like we want to own the data center.

[944] When we say we want to own the data center, we want to own everything in the data center.

[945] and we think arm chips arm cpus can be really a really important part of that arm is not focusing right now enough on that why would they their core market is mobile we want them to do that we think there's a huge opportunity we wanted to own them and and do that and indeed this year invidia announced they are making a data center CPU an arm based data center cp u called grace to go with the new hopper architecture for their latest GPU.

[946] So there's Grace and Hopper.

[947] Of course, the rear admiral, Grace Hopper, I think.

[948] I think that's right.

[949] I think she was in the Navy.

[950] It's a great computer scientist, pioneer.

[951] So yeah, like data center.

[952] It's, it's big.

[953] It's interesting.

[954] So the objectors to that acquisition, and it's a good objection, and this is ultimately, I think, why they abandon it because I get the regulatory pressure on this is arms business is simple.

[955] They make the IP, so you can license one of two things from them.

[956] You can license the instruction set.

[957] So even Apple, who designs their own chips, is licensing the arm instruction set.

[958] And so in order to use that, I don't know what it actually is, 20 keywords or so that can get compiled to assembly language, to run on whatever the chip is, you know, if you want to use these instructions, you have to license it from Arm, great.

[959] And if you don't want to be Apple and you don't want to go build your own chips, or you don't want to be Nvidia or whatever, but you want to use that instruction set, you can also license these off -the -shelf chip designs from us.

[960] And we will never manufacture any of them, but you take one of these two things, you license from us.

[961] You have someone like TSM make them.

[962] Great.

[963] Now you're a fabulous semiconductor company.

[964] And they sell to everyone.

[965] And so of course the regulatory body is going to step in and being like, wait, wait, wait.

[966] So, Nvidia, you're a fabulous chip company.

[967] You're a vertically integrated business model.

[968] Are you going to stop allowing arm licenses to other people?

[969] And Nvidia goes, oh, no, no, no, no. Of course we would never do that.

[970] Over time, they might do some stuff like that.

[971] But the thing that they were sort of like, which is believable, beating the drum on that the strategy was going to be, is right now our whole business strategy is that Kuda, and everything built on top of it, our whole software services ecosystem, is just for our hardware.

[972] And how cool would it be if you could use that stuff on arm designed IP, either just using the ISA or also using the actual designs that people license from them?

[973] How cool would it be if because we were one company, we were able to make all of that stuff available for arm chips as well.

[974] Yeah.

[975] Plausible, interesting, but no surprise at all that they face too much regulatory pressure to go through with this.

[976] No, but clearly that idea rattled around in Jensen's head a bunch and in NVIDIA's head, because, well, let's catch us up to today.

[977] So they just did GTC at the end of March, the big developer, the big GPU developer conference that they do every year that they started in 2009.

[978] as part of building the whole Kuta ecosystem.

[979] I mean, it's so freaking impressive now.

[980] Like, there are now 3 million registered Kuda developers, 450 separate SDKs and models.

[981] For Kuta, they announced 60, 60, new ones at this GTC.

[982] We talk about the next generation GPU architecture with Hopper and then the Grace CPU to go along with it.

[983] I think Hopper, I could be wrong on this.

[984] I think Hopper is going to be the world's first for nanomime.

[985] process chip using TSM's new four nanometer process, which is, I think that's right.

[986] Amazing to talk a lot about Omniverse.

[987] We're going to talk about Omniverse in a second, but you mentioned this licensing thing.

[988] They usually do their investor day, their analyst day, at the same time as GTC.

[989] And in the analyst day, Jensen gets up there.

[990] It's just so funny, I've been going through the whole history of this now of like looking for a market, trying to find some market of any size.

[991] And he's like, we are targeting a trillion dollar market.

[992] He's like a start.

[993] He's like a start up raising a seed round, walk it in with a pitch stick.

[994] We'll put this graphic up on the screen for those watching the video.

[995] It's a articulation of what the segments are of this trillion -dollar addressable opportunity that NVIDIA has in front of it.

[996] My view of this is if their stock price wasn't what it was, there's no way that they would try to be making this claim that they're going after a trillion -dollar market.

[997] I think it's squishy.

[998] Oh, there's a lot a squish in there.

[999] But the fact that they're valued today, I mean, what's their market cap right now?

[1000] Something like half a trillion.

[1001] Half a trillion dollars.

[1002] They need to sort of justify that unless they are willing to have it go down.

[1003] And so they need to come up with a story about how they're going after this ginormous opportunity, which maybe they are, but it leads to things like an investor day presentation of let us tell you about our trillion dollar opportunity ahead.

[1004] And the way that they actually articulate it is we are going to serve customers that represent a hundred trillion dollar opportunity and we will be able to capture about one percent of that.

[1005] God, it's just like a freaking seed company pitch deck.

[1006] If we just get one percent of the market.

[1007] Well, and that's what I think.

[1008] We're going to talk about this in narratives in a minute, but this is a generational company.

[1009] This is unbelievable.

[1010] This is amazing.

[1011] There's so much to admire here.

[1012] This company did what, like 20 -something billion in revenue last year and is worth half a trillion dollars?

[1013] They did $27 billion last year in revenue.

[1014] Google AdWords revenue in the fourth quarter of 2021 was $43 billion.

[1015] Google as a whole did $257 billion in revenue.

[1016] So like, you've got to believe if you're an Nvidia shareholder.

[1017] Right.

[1018] They're the eighth largest company in the world by market cap, but these revenue numbers, you know, are in a different order of magnitude.

[1019] You got to believe it's on the come.

[1020] Yeah, you do.

[1021] I mean, NVIDIA has literally three times the price to sales ratio of Apple or price to revenue as Apple and nearly 2X Microsoft.

[1022] And that's on revenue.

[1023] I mean, fortunately, this NVIDIA story is not speculative in the way that an early stage startup is speculative.

[1024] Like, even if you think it's overvalued, it is still a very cash generative business.

[1025] Yes.

[1026] They generate $8 billion of free cash flow every year.

[1027] So I think they're sitting on $21 billion in.

[1028] cash because the last few years have been very cash generative, very suddenly for them.

[1029] So the takeaway there is by any metric, price of sales, price earnings, all that.

[1030] They're much more richly valued than an Apple or Microsoft or these fang companies.

[1031] But it is, you know, extremely profitable business, even on an operating profits perspective.

[1032] Well, you sell enough with that enterprise data center goodness and you're going to make some money.

[1033] It's crazy.

[1034] They now have a 66 % gross margin.

[1035] So that illustrates to me how seriously differentiated they are and how much of a moat they have versus competitors in order to price with that kind of margin.

[1036] Because think back, we'll put it up on the screen here, but back in 99, they had a gross margin of 30 % on their graphics chips.

[1037] And then in 2014, they broke the 50 % mark.

[1038] And then today, and this slide really illustrates it, it's architecture, systems, data center, Kuda, Kuda, X. Like, it's like the whole stack of stuff that they sell as a solution and is sort of all bundled together.

[1039] And bundle is the right word.

[1040] I think they get great economics because they're bundling so much stuff together.

[1041] It's 66 % gross margin business now.

[1042] Yeah.

[1043] Well, and thinking about increasing that gross margin further and what we were talking about a minute ago with Arm and the licensing.

[1044] So at the analyst day around GTC this year, they say that they're going to start licensing.

[1045] a lot of the software that they make separately licensing it separate from the hardware like kuda and there's a quote from jensen here the important thing about our software is that it's built on top of our platform it means that it activates all of invidia's hardware chips and system platforms and secondarily the software that we do are industry defining software so we've now finally produced a product that an enterprise can license they've been asking for it And the reason for that is because they can't just go to open source and download all the stuff and make it work for their enterprise.

[1046] No more than they could go to Linux, download open source software, and run a multi -billion dollar company with it.

[1047] When we were joking a few minutes ago about you say solution and I see margin, you know, yeah, like open source software companies have become big for this reason, you know, data bricks, confluent, elastic.

[1048] These are big companies with big revenue based on open source because enterprises, they're like, oh, I want that software, but they're not just going to, you know, go to give your JP Morgan.

[1049] You're not going to go to GitHub and be like, great, I got it now.

[1050] You know, you need solutions.

[1051] So to Jensen and Nvidia, they see this as an opportunity to, I'm sure this isn't going to be cannibalizing hardware customers for them.

[1052] I think this is going to be incremental selling on top of what they're already doing.

[1053] That's an important point.

[1054] And I think this is a playbook theme that I had, but oftentimes when someone has...

[1055] hardware that is differentiated by software and services, and then they decide to start selling those software and services a la carte.

[1056] It's a strategy conflict.

[1057] It's your classic vertical versus horizontal problem, unless you are good at segmentation.

[1058] And that's sort of what NVIDIA is doing here, which is what they're saying, well, we're only going to license it to people that there's no way that they would have just bought the hardware and gotten all this stuff for free anyway.

[1059] So if we don't think it's going to cannibalize and there are a completely different segment and we can do things in pricing and distribution channel and terms of service that clearly walls off that segment, then we can behave in a completely different way to that segment.

[1060] Yeah, and get further returns on our assets that we've generated.

[1061] Yep.

[1062] It is a little Tim Cook, though, in, you know, Tim Cook beat in the services narrative drum.

[1063] I mean, it is kind of, you hear public company CEO who has a high market cap and everyone's asking where the next phase of growth is going to come from and saying, we're going to sell services and look at this growing business line of licensing that we have.

[1064] Oh, my goodness.

[1065] But who else is going to do at wearing a leather jacket?

[1066] At is a great point.

[1067] It's a great point.

[1068] Frankly, Elon.

[1069] But, well, we'll talk about cars in a second again.

[1070] Okay.

[1071] So a few other things just to talk about the business today that I think are important to know, just as you sort of like think about, sort of have a mental model for what it.

[1072] Nvidia is.

[1073] It's about 20 ,000 employees.

[1074] We mentioned they did 27 billion in revenue last year.

[1075] We talked about this very high revenue multiple, or earnings multiple, or however you want to frame it relative to fame it, relative to fame companies.

[1076] They're growing much faster than Apple, Microsoft, Google.

[1077] They're growing at 60 % a year.

[1078] This is a 30 -year -old company that grew 60 % in revenue last year.

[1079] If you're not used to, like, wrapping your mind around that, like startups double and triple, but like in the first five years that they exist, Google has had this amazing run where they're still growing at 40%.

[1080] Microsoft went from 10 to 20 % over the last decade.

[1081] Again, amazing.

[1082] They're accelerating, but like, Nvidia is growing as 60%.

[1083] Right.

[1084] I don't care what your discount rate is.

[1085] Having 60 % growth in your DCF model versus 20 or 40 will get you a lot more multiple.

[1086] Inflation be damned.

[1087] Inflation be damned.

[1088] Okay, a couple other things about specific segments of the business that I think are pretty interesting.

[1089] So they have not slept on gaming.

[1090] Like, we keep beating this NVIDIA data center enterprise, machine learning argument.

[1091] Yeah, we haven't even talked about ray tracing and...

[1092] Right, yeah, this RTX set of cards that they came out with.

[1093] The fact that they could do ray tracing in real time, holy crap, for anyone who's looking for sort of a fun dive on how graphics works, go to the Wikipedia page for ray tracing.

[1094] it's very cool.

[1095] You model where all the light sources are coming from, where all the paths would go in 3D.

[1096] The fact that Nvidia can render that in real time at 60 frames a second or whatever while you're playing a video game is nuts.

[1097] And one of the ways that they do that, they invented this new technology that's extremely cool is called DLSS, deep learning super sampling.

[1098] And this I think is like where Nvidia really shines bringing machine learning stuff and gaming stuff together.

[1099] where they basically have faced this problem of, well, we either could render stuff at low resolution with less frames, because we can only render so much per amount of time, or we could render really high resolution stuff with less frames.

[1100] And nobody likes less frames, but everyone likes high resolution.

[1101] So what if we could cheat death, and what if we could get high resolution and high frame rate?

[1102] And they're sitting around thinking, how on earth could we do that?

[1103] And they're like, you know what?

[1104] Maybe this 15 -year bet that we've been making on deep learning can help us out.

[1105] And what they discovered here and invented in DLSS, and AMD does have a competitor to this.

[1106] It's a similar sort of idea.

[1107] But this DLSS concept is totally amazing.

[1108] So what they basically do is they say, well, it's very likely that you can infer what a pixel is going to be based on the pixels around it.

[1109] That's awesome.

[1110] Also pretty likely you can infer what a pixel is going to be based on what it was in the previous frames.

[1111] And so let's actually render it at a slightly lower resolution so we can bump up the frame rate.

[1112] And then when we're outputting it to screen, we will use deep learning to artificially...

[1113] At the final stage of the graphics pipeline.

[1114] Yes.

[1115] Yeah, oh, that's awesome.

[1116] It's really cool.

[1117] And when you watch the side -by -side on all these YouTube videos, it looks amazing.

[1118] I mean, it does involve really tight, embedded development with the game developers.

[1119] They have to sort of do stuff to make it DLSS enabled.

[1120] But it just looks phenomenal.

[1121] And it's so cool that when you're looking at this 4K or even 8K output of a game at, you know, full frame rate, you're like, whoa, in the middle of the graphics pipeline, this was not this resolution.

[1122] And then they magically upscaled it.

[1123] It's basically making the like, enhance joke like a real thing.

[1124] That's so awesome.

[1125] I'm remembering back to the Riva 128 in the beginning of when they went to game developers and they were like, yeah, yeah, yeah, all the blend modes in DirectX, you know, you don't need all them.

[1126] Just use these.

[1127] Yes, exactly, exactly.

[1128] And they have the power to do it.

[1129] I mean, they have the stick and the curate with game developers to do it.

[1130] Oh, I mean, at this point, no game developer is not going to make their games optimized for the latest Nvidia hardware.

[1131] the other thing that is funny that's within the gaming segment, because they didn't want to create a new segment for it, is crypto.

[1132] So because they have poor visibility into it, and before they weren't liking the fact that it was actually reducing the amount of cards that were available to the retail channel for their gamers to go and buy, what they did was they artificially crippled the card to make it worse at crypto mining.

[1133] And then they came out with a dedicated crypto mining card.

[1134] Yes.

[1135] And so like the charitable PR thing from Nvidia is, hey, you know, we really did, we love gamers, and we didn't want to make it so that the gamers couldn't get access to, you know, all the cards they want.

[1136] But really, they're like, hmm, people are just like straight up performing an arbitrage by crypto mining on these cards.

[1137] Let's make that more expensive on the cheap cards and let's make dedicated crypto hardware for them to buy to do those.

[1138] Let's make that our arbitrage.

[1139] Yes.

[1140] Your arbitrage is my opportunity.

[1141] So magically, their revenue is more predictable now and they get to make more money because much like their sort of terms of service data center thing, they terms of serviced their way to being able to create some segmentation and thus more profitability.

[1142] Love it.

[1143] Evil genius laugh.

[1144] The last thing that you should know about Nvidia's gaming segment is this really weird concept of add -in board partners.

[1145] So we've been oversimplifying in this whole episode saying, oh, you know, you go and you buy your RTX -39 DTI at the store, and you run your favorite game on it.

[1146] But actually, you're not buying that from Nvidia, the vast majority of the time.

[1147] You are going to some third -party partner, ASIS, MSI, ZOTAC is one.

[1148] There's also, like, a bunch of really low -end ones as well, who Nvidia sells the cards to, and those people install the cooling and the branding and all this stuff on top of it, and you buy it from them.

[1149] and it's really weird to me that Nvidia does that.

[1150] I love how consumer gaming graphics cards have become the modern day equivalent of a hot rod.

[1151] Oh, dude, as you can imagine for this episode, I've been hanging a lot on the Nvidia subreddit.

[1152] And like, it's not actually about Nvidia or Nvidia the company or Nvidia the strategy.

[1153] It's like, show off your sick photos of your glowing rig, which is pretty funny.

[1154] But like, it feels like a remnant of old Nvidia that they still do this.

[1155] Like, they do make something called the Founders Edition card, and it's basically a reference design where you can buy it from NVIDIA directly, but I don't think the vast majority of their sales actually come from that.

[1156] Oh, it's like, what are the Android phones that Google makes, Pixel?

[1157] Yeah, it's exactly like that, the pixel.

[1158] It's exactly what it is, yeah.

[1159] So I suspect that shifts more over time.

[1160] I can't imagine a company that wants as much control as NVIDIA does loves the ad and board partner thing, but they've built a business on it, and so they're not really willing to cannibalize and alienate.

[1161] But I bet if they had their way and they're becoming a company that can more often have their way, they'll find a way to kind of just go more direct.

[1162] Makes sense.

[1163] Two other things I want to talk about.

[1164] One is automotive.

[1165] So this segment has been like very small from a revenue perspective for a long time and seems to not have a lot of growth.

[1166] But Jensen says in his pitch deck it's going to be a $300 billion part of the dam.

[1167] And I think right now it's something like, is it a billion dollars in revenue?

[1168] I think it's like a billion dollars, but it doesn't really grow.

[1169] I don't even know if it's that much.

[1170] Don't quote me on that.

[1171] So here's what's going on with automotive, which is pretty interesting.

[1172] What Nvidia used to do for automotive is what everyone used to do for automotive, which is make fairly commodity components that automakers buy and then put in there.

[1173] Every technology company has had their fanciful attempt to try to create a meaningfully differentiated experience in the car.

[1174] All have failed.

[1175] You think about Microsoft and the Ford Sync.

[1176] Ford Sync.

[1177] Oh, wow.

[1178] You think about car play, kind of maybe a little bit works.

[1179] And the only company that's really been successful has been Tesla at starting like a completely new car company.

[1180] That's the only way they're able to provide a meaningful, differentiated experience.

[1181] Nvidia is my perception of what they're doing is they're pivoting this business line, this like flat, boring, undifferentiated business line to say maybe EVs, electric vehicles and autonomous driving.

[1182] is a way to break in and create a differentiated experience, even if we're not going to make our own cars.

[1183] And so I think what's really happening here is when you hear them talk about automotive now, and they've got this very fancy name for it, it's the something drive platform.

[1184] Oh, Hyperion Drive, is that it?

[1185] Something like that.

[1186] But dealing with Nvidia's product naming is maddening.

[1187] But this drive platform, it kind of feels like they're making the full EV -A -V hardware software stack except for the metal and glass and wheels and then going to car companies and saying, look, you don't know how to do any of this.

[1188] This thing that you need to make is basically a battery and a bunch of GPUs and cameras on wheels.

[1189] And you're issuing these press releases saying you're going in that direction, but none of this is the core competency of your company except the sales and distribution.

[1190] So like, what can we do here?

[1191] And if Nvidia is successful in this market, it'll basically look like you know, an Nvidia computer, full software, hardware with a car chassis around it that is branded by whatever the car company is.

[1192] Like the Android market.

[1193] Yeah.

[1194] And I think we will see if the shift to autonomous vehicles is a, real, B, near term, and see enough of a dislocation in that market to make it so that someone like Nvidia, a component supplier, actually can get to own a bunch of that value chain versus the auto manufacturer we're kind of forever stubbornly getting to keep all of it and control the experience.

[1195] Yep.

[1196] Which, to do a mini bull and bear on this here before we get to the broader on the company, you know, the bull case for that is we were, again, a friend of the show, Jeremy messaging with in Slack.

[1197] Lotus is one of their partners?

[1198] Is Lotus gonna go build autonomous driving software?

[1199] Like, I don't think so.

[1200] Ferrari?

[1201] No. You know?

[1202] Not at all.

[1203] They're going to be Nvidia cars effectively.

[1204] Okay, last segment thing I want to talk about is how we opened the show, talking about the NVIDIA Omniverse.

[1205] And this is not Omniverse like Metaverse.

[1206] It is similar in that it's kind of a 3D simulation type thing, but it's not an open world that you wander around in the same way that Meta is talking about or that you think about in Fortnite or something like that.

[1207] What they mean by Omniverse is pretty interesting.

[1208] So a good example of it is this Earth 2, this digital twin of Earth that they're creating that has these really sophisticated climate models that they're running, that basically is a proof of concept to show enterprises who want to license this platform, we can do super realistic simulations of anything that's important to you.

[1209] And what their pitch is to the enterprise is, hey, you've got something.

[1210] Let's say it is a bunch of robots that need to do.

[1211] wander around your warehouse to pick and pack if it's Amazon, who actually, Amazon is a customer.

[1212] They showcase Amazon and all their fancy videos, and they say, you're going to be using our hardware and software to train models to figure out the routes for these things that are driving around your data centers.

[1213] You're going to be licensing certainly some of our hardware to actually do the inference to put on the robots that are driving around.

[1214] When you want to make a tweak to a model, you're not just going to like deploy those to all the robots.

[1215] You can, kind of want to run that in the Omniverse first, and then when it's working, then you want to deploy it in the real world.

[1216] And their Omniverse pitch is basically, it's an enterprise solution that you can license from us where anytime you're going to change anything in any of your real world assets, first model it in the Omniverse.

[1217] And I think that's a really powerful, like I believe in the future of that in a big way, because I think now that we have the compute, the ability to gather the data and the ability to actually, you know, run these simulations in a way that has a efficient way of running it and a good user interface to understand the data.

[1218] People are going to stop testing in production with real world assets and everything's going to be modeled in the Omniverse first before rolling out.

[1219] This is what an enterprise metaverse is going to be.

[1220] This is not designed for humans.

[1221] Humans may interact with this.

[1222] There will be UI.

[1223] You will be able to be part of it.

[1224] The purpose of this is for.

[1225] simulating applications, and most of it, I think, is going to run with no humans there.

[1226] Yep, pretty crazy.

[1227] Yeah, it's a good idea.

[1228] Sounds like a good idea.

[1229] All right.

[1230] You want to talk Barron Bull case on the company?

[1231] Let's do it.

[1232] Analysis.

[1233] So, I mean, they paint the bullcase for us when they say there's a $100 trillion future.

[1234] We're going to capture 1 % of it.

[1235] There's $300 billion from automotive.

[1236] Here's the four or five segments that add up to a trillion dollars of opportunity.

[1237] Sure.

[1238] That's like a very neat way with a bow on it and a very wishy -washy, hand -wavy way of articulating it.

[1239] So the question sort of becomes, where's AMD fall in all this?

[1240] They're a legitimate second -place competitor for high -end gaming graphics, and I think we'll continue to be.

[1241] That feels like a place where these two are going to keep going head -to -head.

[1242] The bare case is that there's a TikTok rather than a durable competitive advantage for NVIDIA.

[1243] But most high -end games you can play on both AMD and NVIDIA hardware.

[1244] at this point.

[1245] The question for the data center is, is the future these general purpose GPUs that Nvidia continues to modify the definition of GPU to include specialized functions as well, all this other stuff they're putting in their hardware, or is there someone else who is coming along with a completely different approach to accelerated computing, whether accelerating workloads off the GPU onto something new, like a cerebris or like a graphcore that is going to eat their lunch in the enterprise AI data center market?

[1246] That's an open question.

[1247] You know, it's interesting.

[1248] Like, people have been talking about that for a while.

[1249] The other big bear case that people have been talking about, again, for a while now, is, you know, the big, big customers of Invidia.

[1250] that are paying them a lot of money.

[1251] The Teslas, the Googles, the Facebooks, the Amazon's, the Apples.

[1252] And not just paying them a lot of money and getting, you know, assets of value of that.

[1253] They're paying high gross margin dollars to Invidia for what they're getting.

[1254] That those companies are going to want to say, you know, it's not that hard to design our own silicon to bring all this stuff in -house.

[1255] We can tune it to exactly our use cases, sort of similar to the cerebris, uh, GraphCore bear case on NVIDIA.

[1256] I think in both of these cases, you know, it hasn't happened yet.

[1257] Well, there have been a lot of people who have made a lot of noise, but there have been few that have executed on it.

[1258] Like, Apple has their own GPUs on the M1s, Tesla's switching.

[1259] Hasn't happened yet, but switching to their own, for the full self -driving, they're doing their own stuff on the car, and they're switching.

[1260] Yep, that is switch.

[1261] On the inference side.

[1262] Yes.

[1263] On device, yes, that has happened.

[1264] But look, Nvidia is probably strong in that, but I think the real thing to watch is, the data center.

[1265] And Google is probably the biggest bear case there.

[1266] It's interesting to talk about these companies, and particularly Cerebrus, because what they're doing is such a gigantic swing and a totally different take than what everyone else has done.

[1267] For folks who hasn't sort of followed the company, they're making a chip that's the size of a dinner plate.

[1268] Everyone else's chip is like a thumbnail, but they're making a dinner plate size chip.

[1269] And you know, the yields on these things kind of suck.

[1270] So, like, they need all the redundancy on those huge chips to make it so that...

[1271] Oh, my God, the amount of expense to do that.

[1272] Right.

[1273] And you can put one on a wafer.

[1274] Oh.

[1275] These wafers are crazy expensive to make.

[1276] Wow.

[1277] So you get poor yields in the wrong places on a wafer, and, like, that whole wafer is toast.

[1278] Right.

[1279] So a big part of the design of Cerebrus is this sort of redundancy and the ability to turn off different pieces that aren't working.

[1280] They draw 60 times as much power.

[1281] they're way more expensive.

[1282] Like if if Nvidia is going to sell you a $20 ,000 or $30 ,000 ship, Cerebrus is going to sell you a $2 million ship to do AI training.

[1283] And so it is this bet in a big way on hyper -specialized hardware for enterprises that want to do these very specific AI workloads.

[1284] And it's deployed in these beta sites in research labs right now.

[1285] And, you know, not there yet, but it'll be very interesting to watch if they're able to meaningfully compete for what everyone thinks will be a very large market, these enterprise AI workloads.

[1286] I mentioned Google that made a bunch of noise about making their own silicon in the data center and then stayed the course and stayed really serious about it with their TPUs.

[1287] Their business model is different.

[1288] So nobody knows what the bill of materials is to create a TPU.

[1289] Nobody knows really what they cost to run.

[1290] They don't retail them.

[1291] They're only available in Google Cloud.

[1292] And so Google is sort of counterpositioned against Nvidia here, where they're saying, we want to differentiate Google Cloud with this offering that depending on your workload, it might be much cheaper for you to use TPUs with us than for you to use Nvidia hardware with us or anyone else.

[1293] And they're probably willing to eat margin on that in order to grow Google Cloud's share in the cloud market.

[1294] Interesting.

[1295] So it's kind of the Android strategy, but run in the data center.

[1296] One thing we haven't mentioned, but we should, is cloud is also part of the NVIDIA story, too.

[1297] Like, you can get NVIDIA GPUs in AWS and Azure and Google Cloud, and that is part of the growth story for NVIDIA too.

[1298] And NVIDIA is starting their own cloud.

[1299] You can get direct from NVIDIA cloud -based GPU.

[1300] Data center.

[1301] GPUs, interesting.

[1302] Yeah.

[1303] And it'll be very interesting to see how this all shakes out with the NVIDIA, the startups, and with Google.

[1304] I mean, all that said, though, like, I think, but look, NVIDIA's very, very, very, very richly valued on a valuation basis right now, very, with another very in there.

[1305] It depends if you think their growth will continue.

[1306] Are they a 60 % growing company year over year over year for a while?

[1307] Then they're not richly valued.

[1308] But if you think it's a COVID hiccup or a crypto hiccup.

[1309] But to the bull bear case and kind of both the startups and the big tech company is doing this stuff in -house.

[1310] it's not so easy, you know, like, yeah, Facebook and Tesla and Google and Amazon and Apple are capable of doing a lot, but we just told this whole story.

[1311] This is 15 years of Kuda and the hardware underneath it and the libraries on top of it that Nvidia has built to go recreate that and surpass it on your own is such an enormous, enormous bite to bite.

[1312] Yes.

[1313] And if you're not a horizontal player and you're a vertical player, you better believe that the pot of gold at the end is worth it for you for this massive amount of cost to create what NVIDIA has created.

[1314] Yep.

[1315] Like, Nvidia has the benefit of getting to serve every customer.

[1316] If you're Google, and their strategy is what I think it is, of not retailing TPUs at any point, then your customer is only yourself so you're constrained by the amount of people you can get to use Google Cloud.

[1317] Well, and at least with Google, they have Google Cloud that they can sell it through.

[1318] Yep.

[1319] Power.

[1320] Ooh, power.

[1321] So the way I want to do this section, because in our NVIDIA episode we covered the first 13 years of the company, we talked a lot about what does their power look like up to 2006.

[1322] And now I want to talk about what does their power look like today.

[1323] What is the thing that they have that enables them to have a sustainable competitive advantage and continue to maintain pricing power over their nearest competitor, be it Google Cerebrus in the enterprise, or AMD in gaming.

[1324] Yep.

[1325] And just to enumerate the powers again, as we always do, counter positioning, scale economies, switching costs, network economies, process power, branding, and cornered resource.

[1326] So there are definitely scale economies.

[1327] The whole Kuta investment...

[1328] Yes.

[1329] Not at first, but definitely now, is predicated on being able to amortize that 1 ,000 -plus employees spend over the base of the 3 million developers and all the people who are buying the hardware to use what those developers create.

[1330] This is the whole reason we spent 20 minutes talking about if you were going to run this playbook, you needed an enormous market to justify the capex you were going to put in.

[1331] Right.

[1332] So very few other players have access to the capital and the market that Nvidia does to make this type of investment.

[1333] So they're basically just competing against AMD for this.

[1334] Totally agree.

[1335] Scale economies to me is like the biggest one that pops out, to the extent that you have lock -in to developing on Kuda, which I think a lot of people really have lock -in on Kuda, then that's major switching costs.

[1336] Yep.

[1337] Like, if you're going to boot out NVIDA, that means you're booting out Kuda.

[1338] Is Kuda a cornered resource?

[1339] Oh, interesting.

[1340] Maybe, I mean, it only works with NVIDIA hardware.

[1341] You could probably make an argument there's process power, or at least there was somewhere along the way with them having the six -month ship cycle advantage that probably has gone away since people trade around the industry a lot and that wasn't sort of a hard thing for other companies to figure out.

[1342] Yeah, I think process power definitely was part of the first instantiation of Nvidia's power to the extent it had power.

[1343] Right.

[1344] Yeah, I don't know as much today, especially because TSM will work with anybody.

[1345] In fact, TSM is working with these new startup billion -dollar funded silicon companies.

[1346] Yes, they are.

[1347] Yes.

[1348] Yeah, it's funny.

[1349] I actually heard a rumor, and we can link to it in the show notes, that the Ampere series of chips, which is the one immediately before the hopper, the sort of A -series chips, are actually fabbed by Samsung, who gave them a sweetheart deal.

[1350] NVIDIA likes to keep the lore alive around TSM, because they've been this great long -time partner and stuff, but they do play manufacturers off each other.

[1351] I even think that Jensen said something recently like Intel has approached us about fabbing some of our chips and we are open to the conversation.

[1352] Yes, yes, that did happen.

[1353] So there was this big cybersecurity hack a couple of months ago by this group Lapsis and they stole access to NVIDIA's source code.

[1354] And actually Jensen went on Yahoo Finance and talked about the fact that this happened.

[1355] I mean, this is a very public incident.

[1356] And it's clear from the demands of Lapsis where some of NVIDIA's power lies, because they demanded two things.

[1357] They said, one, get rid of the crypto governors, like make it so that we can mine, which may have been a red herring that might have just been them trying to look like a bunch of like crypto miner people.

[1358] And the other thing they demanded is that NVIDIA open source all of its drivers and make available its source code.

[1359] I don't think it was for Kuda.

[1360] I think it was just the drivers, but it was very clear that, like, we want you to make open your trade secrets so that other people can build similar things.

[1361] And that, to me, is illustrative of the incredible value and pricing power that Nvidia gets by owning not only the driver stack, but, you know, all of Kuda and how tightly coupled their hardware and software is.

[1362] NVIDIA is we just did this Armbrust recent episode with Hamilton and Chen Yi.

[1363] NVIDIA is a platform in my mind.

[1364] No doubt about it.

[1365] Kuda and NVIDIA and general purpose computing on GPUs as a platform.

[1366] So whatever, you know, all of the stew of powers that go into making that go into making Apple, Microsoft, you know, and the like, go into NVIDIA.

[1367] Yep.

[1368] I think the stew of powers is the right way to phrase that.

[1369] Yes.

[1370] Anything else here?

[1371] You want to move to Playbook?

[1372] Let's move to Playbook.

[1373] So, man, I have, I just wrote down in advance, one that is such a big one for me. And I'm biased because I try to think about this in investing, particularly in public markets investing.

[1374] But like, man, you really, really want to invest in whoever is selling the picks and the shovels in a gold rush.

[1375] The AI, you know, ML, deep learning gold rush.

[1376] Those years, gosh, oh my gosh.

[1377] Like, we should all, all be kicking ourselves of 2012.

[1378] maybe not 2012, but certainly 2014, 2015 into 2016, like, duh, you know, Mark Andresen saying every startup that comes in here that wants to do AI and deep learning and they're all using Nvidia, like maybe we should have bought Nvidia.

[1379] Like, I don't know if any one of those startups, any given one is going to succeed, but I'm pretty sure in video is going to succeed back then.

[1380] Yeah, it's such a good point.

[1381] Kicking myself.

[1382] One I have is being willing to expand your mission.

[1383] So it's funny how Jensen, early days, would talk about to enable graphics to be a storytelling medium.

[1384] And of course, this led to the invention of the pixel shader and the idea that everybody can sort of tell their own visual story, their own way, in a social networked real -time way, very cool.

[1385] And now it's much more that wherever there is a CPU, there is an opportunity to accelerate that CPU.

[1386] And Nvidia will bring accelerated computing to everyone.

[1387] And we will make all the best hardware, software, and services solutions to make it so that any computing workload runs in the most efficient way possible through accelerated computing.

[1388] That's pretty different that enable graphics as a storytelling medium.

[1389] But also, they need to sell a pretty big story around the Tam that they're going after.

[1390] I think there's also something to the whole NVIDIA story, you know, across the whole arc of the company of, you know, it's sort of a trite cliche thing at this point in startup land, but so few companies and founders can actually do it.

[1391] Just not dying.

[1392] Yeah.

[1393] They should have died at least four separate times.

[1394] And they didn't.

[1395] And part of that was brilliant strategy.

[1396] Part of that was things going their way.

[1397] But I think a large part of it too was just the company and Jensen, particularly in these most recent chapters where there are already a public company just being like, yeah, I'm willing to just sit here and endure this pain.

[1398] And I have confidence that, like, we will figure it out.

[1399] The market will come.

[1400] Not going to declare game over.

[1401] One that I have is, we mentioned at the top of the show, but the scale of everything involved in machine learning at this point, and anything semiconductors, is kind of unfathomable.

[1402] You and I mentioned falling down the YouTube rabbit hole with that Asianometry channel, and I was watching a bunch of stuff on how they make the Silicon Wafers And my god, floor planning is this just unbelievable exercise at this point in history, especially with the way that they sort of overlay different designs on top of each other on different layers of the chip.

[1403] Yeah.

[1404] See more about what floor planning is.

[1405] I bet a lot of listeners won't know.

[1406] So it's funny how they keep appropriating these sort of real world large scale analogies to chips.

[1407] So floor planning, the way that an architect would lay out the 15 rooms in a house or five rooms in a house or two rooms in a house.

[1408] on a chip is laying out all of the circuitry and wires on the actual chip itself, except, of course, there's like 10 million rooms.

[1409] And so it's incredibly complex.

[1410] And the stat that I was going to bring up, which was just mind bending to think about, is that there are dozens of miles of wiring on a GPU.

[1411] Wow.

[1412] That is mind bending.

[1413] Because these things are like, you know, I don't know, they're less than the size of your palm, right?

[1414] Right.

[1415] And it obviously is not wiring in the way you think about like a wire.

[1416] I'm going to reach down and pick up my Ethernet cable, but it's wiring in the EUV etched substrate on -ship exposure is probably the term that I'm looking for here, photolithography exposure.

[1417] But it is just so tiny.

[1418] I mean, you can say four nanometers all you want, David, but that won't register with me how freaking tiny that is until you're sort of faced with the reality of dozens of miles of quote -unquote wires on this chip.

[1419] Yeah, it's not like, to me, that registers as like, oh, yeah, that's like a decal I put on my hot rod.

[1420] Four nanometers.

[1421] I got the S version.

[1422] But yeah, like, that's what that means.

[1423] Okay, here's one that I had that we actually even talked about, which I think will be fun.

[1424] So I generated a CAPX graph.

[1425] Ooh, fun.

[1426] We'll show it on screen here for those watching on video.

[1427] Obviously, there's a very high looking line for Amazon because building data centers and fulfillment centers is very expensive, especially in the last couple of years when they're doing this massive build -out.

[1428] But imagine without that line for a minute.

[1429] Nvidia only has a billion dollars of CAPEX per year.

[1430] And this is relative for people listening on audio relative to a bunch of other, you know, Fang -type companies.

[1431] Yeah, so Apple has $10 billion of spend on capital expenditures per year.

[1432] Microsoft and Google have $25 billion.

[1433] TSM, who makes the chips, has $30 billion.

[1434] what a great capital -efficient business that NVIDIA has on their hands only spending a billion dollars a year in Cappex.

[1435] It's like it's a software business.

[1436] And it basically is...

[1437] Well, it is, right?

[1438] Like TSM does the fabbing.

[1439] NVIDIA makes software and IP.

[1440] Yep.

[1441] So here, this is the best graph for you to very clearly see the magic of the fabless business model that Morris Chang was so gracious to invent when he grew TSM.

[1442] Thank you, Morris.

[1443] another one that I wanted to point out, it's a freaking hardware company.

[1444] I know they're not a hardware company, but they're a hardware company with 37 % operating margins.

[1445] So this is even better than Apple.

[1446] And for non -finance folks, operating margins, so we talked about their 66 % gross margin.

[1447] That's like unit economics.

[1448] But that doesn't account for all the headcount and the leases and just all the fixed costs in running the business.

[1449] Even after you subtract all that out, 37 % of every dollar that comes in gets to be kept by Nvidia shareholders.

[1450] It's a really, really, really cash -generative business.

[1451] And so if they can continue to scale and can keep these operating margins or even improve them because they think they can improve them, that's really impressive.

[1452] Wow.

[1453] I didn't realize that's better than apples.

[1454] Yeah.

[1455] I think it's not as good as like Facebook and Google because they just run these like...

[1456] Well, those are digital monopolies.

[1457] Like, come on.

[1458] Basically zero -cost digital monopolies in some of the largest markets in history.

[1459] But it's still very good.

[1460] Our sponsor for this episode is a brand -new one for us.

[1461] Statsig.

[1462] So many of you reached out to them after hearing their CEO, Vijay, on ACQ2, that we are partnering with them as a sponsor of Acquired.

[1463] Yeah, for those of you who haven't listened, Vijay's story is amazing.

[1464] Before founding Statsig, Vijay spent 10 years at Facebook where he led the development of their mobile app ad product, which, as you all know, went on to become a huge part of their business.

[1465] He also had a front row seat to all of the incredible product engineering tools that let Facebook continuously experiment and roll out product features to billions of users around the world.

[1466] Yep.

[1467] So now Statsig is the modern version of that promise and available to all companies building great products.

[1468] Statsig is a feature management and experimentation platform that helps product teams ship faster, automate AB testing, and see the impact every feature is having on the core business metrics.

[1469] The tool gives visualizations backed by a powerful stats engine unlocking real -time product observability.

[1470] So what does that actually mean?

[1471] It lets you tie a new feature that you just shipped to a core metric in your business and then instantly know if it made a difference or not in how your customers use your product.

[1472] It's super cool.

[1473] Statsig lets you make actual data -driven decisions.

[1474] about product changes, test them with different user groups around the world, and get statistically accurate reporting on the impact.

[1475] Customers include Notion, Brex, OpenAI, FlipCart, Figma, Microsoft, and Cruise Automation.

[1476] There are, like, so many more that we could name.

[1477] I mean, I'm looking at the list, Plex and Vercel, friends of the show at Rec Room, Vanta.

[1478] They, like, literally have hundreds of customers now.

[1479] Also, Statsig is a great platform for rolling out and testing AI product features.

[1480] So for anyone who's used Notion's awesome generative AI features and watched how fast that product has evolved, all of that was managed with Statsig.

[1481] Yep.

[1482] If you're experimenting with new AI features for your product and you want to know if it's really making a difference for your KPI's stat sig is awesome for that.

[1483] They can now ingest data from data warehouses.

[1484] So it works with your company's data wherever it's stored so you can quickly get started no matter how your feature flagging is set up today.

[1485] You don't even have to migrate from any current solution.

[1486] might have.

[1487] We're pumped to be working with them.

[1488] You can click the link in the show notes or go on over to stat sig .com to get started.

[1489] And when you do, just tell them that you heard about them from Ben and David here on Acquired.

[1490] Okay.

[1491] Grating.

[1492] So I think the way to do this one, David, is what's the A plus case?

[1493] What's the C case?

[1494] What's the F case?

[1495] I think so.

[1496] And there's sort of an interesting way to do this one because you could do it from a shareholder perspective where you have to evaluate it based on where it's trading today and sort of like what needs to be true in order to have a a plus investment starting today that sort of thing you mean like a michael mobson expectations investing style yes exactly or you could sort of close your eyes to the price and say let's just look at the company if you're jensen what do you feel would be an a plus scenario for the company regardless of the investment case i kind of think you have to do the one, though.

[1497] Like, I kind of think it's a cop -out to not think about it, like, what's the bull and bear investment case from here?

[1498] As we pointed out, many times on the episode, there's a lot you got to believe to be a bull on Nvidia at this share price.

[1499] So what are they?

[1500] Well, one big one is that they continue their incredible dominance and they're, what are they growing, like 75 % or something year over year in the data center?

[1501] Yep.

[1502] and they just sort of continue to own that market.

[1503] I think there's a plausible story there around all the crazy gross margin expansion they've had from sort of selling solutions rather than fitting into someone else's stuff.

[1504] I also think with the Melanox acquisition, there's a very plausible story around this idea of a data processing unit and around being your one -stop shop for AI Data Center hardware.

[1505] And I think rather than saying, like, oh, the upstart competition will fail.

[1506] I think you kind of have to say that NVIDIA will find a way to learn from them and then integrate it into their strategy too.

[1507] Which seems plausible.

[1508] Yeah, but they've been very good at changing the definition of GPU over time to mean more and more robust stuff and accelerate more and more compute workloads.

[1509] And I think you just have to kind of bet that because they have the developer attention, because they now have the relationships to sell into the enterprise, they're just going to continue to be able to do their own innovation, but also fast follow when it makes sense to redefine GPU as something a little bit heftier and incorporate other pieces of hardware to do other workloads into it.

[1510] Yep.

[1511] I think the question for me on an A -plus outcome for Nvidia from this shareholder perspective is, do you need to believe that all the real -world AI use cases are going to happen.

[1512] Do you need to believe that some basket, maybe not all of them, but that some basket of autonomous vehicles, the omniverse, robotics, one or multiple of those three are going to happen.

[1513] They're going to be enormous markets and that Nvidia is going to be a key player in them.

[1514] I mean, I think you do because I think that's where all the data center revenue is coming from is companies that are going after those opportunities.

[1515] I'm wrestling with whether that is something you have to believe or whether that's optionality.

[1516] The reason it would be only optionality, only upside, is if the digital AI, we know that that's a big market.

[1517] There's no question about that at this point.

[1518] Is that going to continue to just get so big?

[1519] Are we still only scratching the surface there?

[1520] How much more AI is going to be baked into all the stuff we do in the digital world?

[1521] And will Nvidia continue to be at the center of that?

[1522] I don't know.

[1523] I don't a great way to assess how much growth is left there that is kind of the right question though yeah they're in an interesting point right now you know there was all the early company stuff that we talked about in the first episode but at the beginning of this episode you know jensen was really asking you to believe it's like hey we're building this kuta thing just ignore that there's no real use case for it or market now there is a real real use case in market for it which is machine learning, deep learning in the digital world, undeniable.

[1524] He's also pitching now that that will exist in the physical world, too.

[1525] Yeah, the A plus is definitely that it does exist in the physical world, and they are the dominant provider of everything you need to be able to accomplish that.

[1526] Yep.

[1527] And if the real world stuff, you know, these little robots that run around factory floors and the autonomous vehicles, and if that stuff doesn't materialize, then, yeah, there's no way that it can support the growth that it's been on.

[1528] I think that's probably right.

[1529] That would be my hunch.

[1530] Although, saying that, though, does feel like a little bit of a betting against the internet.

[1531] You know, like, I don't know, man. Digital world's pretty big and it keeps getting bigger.

[1532] Yeah, but I think we're saying the same thing.

[1533] I think you're saying that these physical experiences will become more and more intertwined with your digital experiences.

[1534] Yeah.

[1535] Yeah.

[1536] I mean, autonomous driving in electric vehicles is an individual.

[1537] internet bet.

[1538] In part, if you want to bet on the growth of the internet, it'll mean you'll drive less.

[1539] But it also means that you're just going to be on the internet when you're driving.

[1540] Yep.

[1541] Yeah.

[1542] Or when you're in motion in the physical world.

[1543] That's actually, that's a bullcase for Facebook, right?

[1544] Is autonomous vehicles, because if people are being driven instead of driving, that's more time they're on Instagram.

[1545] Right.

[1546] That's so true.

[1547] Okay, what's the failure case?

[1548] It's actually quite hard to imagine a failure case of the business in any short order.

[1549] It's very easy to imagine a failure case for the stock in short order if there's a cascading set of events of people losing faith.

[1550] I think maybe the failure case is this amazing growth for the past couple of years was pandemic pull forward.

[1551] It's so hard for me to imagine that that's like to the degree of a Peloton or a Zoom or something like that.

[1552] Right.

[1553] By the way, I think a great company.

[1554] They just got everything pulled forward.

[1555] I don't think Nvidia got everything pulled forward.

[1556] They probably got a decent amount pulled forward.

[1557] Hard to quantify, hard to know, but it's the right thing to be thinking about.

[1558] Yeah.

[1559] All right, carve outs.

[1560] Ooh, carve outs.

[1561] I've got a fun one, small one, well, a collection of small things.

[1562] Longtime listeners probably know one of my favorite, I think my favorite series of books that have been written in the past 10 years.

[1563] years is the expanse series amazing sci -fi nine books so great the ninth book came out last fall it was just even with like a newborn I made time to read this book that's awesome newborn plus acquired I was like I got to read that you know that's how you know recently last month so the authors have been writing short stories like companion short stories alongside the main narrative over the last decade that they've been doing this and they really released a compendium of all the short stories, plus a few new ones, called Memories Legion.

[1564] And it's this really cool.

[1565] Like, I mean, they're great writers, great short stories to read, even if you don't know anything about the expanse story.

[1566] But if you know the whole nine book saga and then these like just paint little, give you a little glimpses into corners and like characters that just exist and you don't question otherwise, but you're like, oh, what's the backstory of that?

[1567] I've been really enjoying that.

[1568] So it's like the solo of the fantastic beast and where to find them.

[1569] Exactly.

[1570] It's like nine or ten of those.

[1571] mirrorless plus big long zoom lens person, but it's kind of annoying to lug that around.

[1572] And then once I started downgrading my phone from the massive, awesome iPhone with the 3X zoom, and I now have the iPhone 13 mini, I think that's what it is, with the two cameras and no zoom lens is really disappointing.

[1573] So it's pretty awesome.

[1574] It fills a sort of spot in my camera lineup to have a point and shoot with a really long zoom lens on it.

[1575] And of course, like, it's not as nice as having a, you know, full frame mirrorless with like an actual zoom lens, but it really gets the job done.

[1576] And it's nice to have that sort of like real feeling mirrorless style image that is very clearly from a real camera and not from a phone that is, uh, it's slightly more inconvenient to carry because you kind of need another pocket.

[1577] Yeah, I was going to ask, can you put it in your pocket?

[1578] Yeah, I put it in a pocket.

[1579] I don't have to have a sort of like a rapid strap around my neck, which is nice.

[1580] Nice.

[1581] So the Sony Arc 100, great little device.

[1582] It's like the seventh generation of it.

[1583] And they've really refined the industrial design at this point.

[1584] That's awesome.

[1585] That's awesome.

[1586] I actually just bought my first camera cube, like a travel camera cube thing for our Alpha 7Cs now that we have, literally it's four acquired for when after the Altimeter episode, I was like, oh, wow.

[1587] We've got to do more in person.

[1588] Yeah.

[1589] Yeah, Ben brought his down and it's like, for sure, I'm going to need to bring this somewhere.

[1590] These cameras are just, they are so good.

[1591] They're so good.

[1592] All right, listeners, thank you so much for listening.

[1593] You should come chat about this episode with us in the Slack.

[1594] There's 11 ,000 other smart members of the acquired community, just like you.

[1595] And if you want more acquired content after this and you were all cut up, go check out our LP show by searching Acquired LP show in any podcast.

[1596] player.

[1597] Hear us interview Nick and Lauren from Trova Trip most recently.

[1598] And we have a job board.

[1599] Acquired .fm slash jobs.

[1600] Find your dream job curated just by us, the fine folks at the Acquired podcast.

[1601] And we will see you next time.

[1602] We'll see you next time.