If you have a 401k or you own the SP or the Nasdaq or straight out NVIDIA stock, then what I'm about to show you could save you or cost you thousands, if not hundreds of thousands of dollars in your retirement. Nvidia's seemingly unstoppable dominance in AI chips is being challenged by the most unexpected competitors, their own biggest customers. And this threat is bigger than most investors realize right now. We're in the denial phase. We're literally talking about 40% of Nvidia's revenue coming from just two customers who are now building their own chips to replace NVIDIA's. And this could trigger one of the biggest valuation resets in tech stock history. My promise to you is that by the end of this video, you know exactly what's happening, when the market will start pricing this in, and what you need to do with your NVIDIA shares before it's too late. This is not an NVIDIA doom and gloom video, but there is a amount of realism we need to take on board because a lot of people have fallen in love with a stock that's made them a lot of money. My name is Felix Preen, I'm an ex-investment banker. That back there was Winston, who's done all the research for this. And we've also founded the GOAT Academy because we're retired and we wanted to spread some joy in education, and we've taught over 20,000 students so far. I'm also the co-founder of Tradevision.io, where we provide Wall Street great market intelligence to regular investors like you and me. So if you want to know what's going on with NVIDIA, the very thing I would do right now, it's Black Friday. It's never been better, it's never been cheaper. You want to add NVIDIA to your news alerts that you're gonna get. And then you're gonna get updates like this, like the news up there on Oracle or something like that. And you're gonna get that live as it happens when it happens. And all you then do is you tap on the news item, it loads the app up for you, and you can see I tapped on a Robin Hood news alerto. You see exactly what's going on there, and we only give you the news that is actually gonna move the stock. Right? We don't give you all the noise and the bollocks and the mainstream media politics because we're only here because we want to make more money for ourselves and our families and golden retrievers, obviously, you know, likes to eat a lot. And that's our mission here. Now, Wall Street is already watching this very closely, and I believe that you need to know about it. Not when it's already priced into the stock, but right now. So I'm gonna share my presentation for you with you because that way you get the most information. I'm also gonna give you for free a workbook to make sure this actually lands for you because this is how important this is. The workbook's free. You can get that in our free community. More than 10,000 people joined out in the last three or four weeks. Go to Phoenix Ranson Oxlash Resource and download the workbook. We all know Nvidia is the king of AI, undisputed, right? They made$51 billion just last quarter in data center revenue alone. That's more revenue in three months than most Fortune 500 companies make in the entire year. The stock's been an insane winner. It's driven the entire market up. And for many Americans, particularly, NVIDIA is one of the largest holdings in their retirement account directly or through index funds, right? Now, how did NVIDIA become so essential? Every major AI breakthrough, ChatGPT, Google's Gemini, Meta's Llama, they've all been trained on NVIDIA's GPUs. Actually, Meta's Lama has apparently been trained on Zuckerberg's Lamas, true story. But the NVIDIA chips became the picks and the shovels of this AI gold rush with them, which isn't over, by the way. It's just shifting. But most investors don't understand it yet. Their success has created a problem. The very companies buying hundreds of billions of dollars worth of NVIDIA chips are now building their own chips. And they have the money and the motivation to succeed. In 2025 alone, tech companies think Microsoft, Amazon, Google, Meta, and Oracle as spending over$380 billion in one year. And a huge chunk of that money is going to NVIDIA. But these companies are looking at their builds and they're going, why are we making NVIDIA so profitable? Why are we making NVIDIA shareholders so wealthy when we could just build our own chips? And that's exactly what they're doing. Nvidia's Achilles heel is the following 39% of their revenue comes from just two customers. Not made up data. Real data from NVIDIA's own presentation. So almost 40% of their revenue comes from two customers. One customer, customer A, has 23% of total revenue, and another 16% of revenue one customer. Now NVIDIA doesn't name these customers, but everybody in the industry overwhelmingly believes that these two are Microsoft and Amazon, biggest cloud companies out there. So why is this dangerous? Well, think about this from a business point of view. If you run a company and 40% of all your money came from just two customers, what would happen if one of them decided to buy, say, 30% less? Your stock would crater. And this isn't theoretical. This is how Wall Street will react when these customers start shifting to their own chips. These customers have immense negotiating power over NVIDIA's price, you would think. But in reality, they don't, because Nvidia says we're fully sold out. Take it or leave it. But these are not typical customers. These are customers who can actually become competitors. And they have an insane cost motivation. When you buy an Nvidia GPU, you are paying a 75% gross margin. What does that mean? It means if you spend$100 on NVIDIA chips, you know, because it's public information, that NVIDIA makes a$75 profit on that. So you're thinking, I could get like this only costs$25. Why am I paying four times more than what this actually costs to build? And now imagine your Microsoft or Google and you're spending tens of billions of dollars on this. You've got the smartest people in the world working for you, and you're thinking, what if we just designed our own chips? We could get rid of that 75% profit margin. We could optimize the chips for our specific tasks, workloads as they call them, and we might even be able to get a competitive advantage because we can do this way, way, way cheaper, right? First thing I want to say to you is download the free workbooks. You really understand what I'm talking to about here. Second is stay abreast of this news story. This is gonna be the news story of 2026 and 2027. But if you know this first, you set up your news alerts on TradeVision for these companies. And I'm gonna give you not just the big tech names, but also their supply chain. Because that's where the real opportunity lies to make some money out of this. You're gonna be ahead of everybody else. So go to Felix Renzo.org slash TradeVision. We have an insane Black Friday offer. It's always crazily cheap anyway. And I think it's literally 11 bucks or something. Something crazy the Black Friday thing. So take advantage of that. And you will ping, get these things pinged up on your phone, and you will not be left in the lurch like everybody else. He's gonna find out about this after it happens, right? So we've got Google. Now Google has for decades been developing its own custom AI chips. It's called Tensor Processing Units, TPUs. And I know you learned CPUs in the 90s, and then you tried to figure out what GPUs were, and now they are TPUs. Give us a frickin' break. But this literally is something that started in 2013 when they realized that if every Android user used voice search for just three minutes a day, they would need double their entire global data center capacity. That's how expensive would be to run an AI workload on a normal chip. So Google did something fairly radical for a software company. They designed their own silicon from scratch. Now, I'm not a techie, but here's the key difference in simple terms. Nvidia's GPUs are all-purpose processors. They're really flexible. They can do many different things. I've got one on this computer, I've got one in my laptop, they're amazing. People are gonna keep buying them. I'm not saying they're gonna stop buying them completely, but Google's TPUs are specialists. They're designed to do one thing extremely well. And that is the maths that powers AI. So think of it like this: GPU is like your Swiss Army knife. It can do many things. It's got a screwdriver, it's got all sorts of stuff in it. But the TPU is a professional chef's knife. It's designed for one specific cut, and it's better at that task than any Swiss knife could ever be. And TPUs can literally be four to ten times more cost effective than Nvidia's GPUs. So let me say that again, up to 10 times cheaper for the same work. They are also two to three times more power efficient, which means they use less electricity. And if you've been paying attention, the one thing that's holding back Google and Microsoft and Meta and Oracle is availability of electricity. There isn't enough of this stuff. So when you're running massive data centers, electricity isn't just your biggest cost, it's also your biggest growth constraint. And Google has just put out something called TPU version 7. It performs, it's called ironwood. It delivers 100% better performance per watt, so per energy unit, than the previous generation. And former Google employees have confirmed these advantages. One former Google executive said the version 6 of this, we're now at version 7, that was 60% more efficient than GPUs for AI. There is a client who remains unnamed, used both Nvidia's H100s and Google's TPUs, and he said, if I were to use eight H100s versus one TPU, I'd spend a lot less money on the one TPU. But the real-world proof is this: Google has trained its Gemini 3 model, which is insanely good, by the way, TriAm. It's one of the most powerful AI models in the world, entirely on TPUs. They did not use any NVIDIA chips. So it shows that TPUs aren't just good for some small tasks, they can handle the most demanding training workloads. And even Jensen Huang, Nvidia's CEO, has acknowledged Google's TPU has a special case among custom chips. And there is even a rumor out there that Jensen called Sam Altman when he heard OpenAI might rent some Google TPUs, which shows you how seriously Nvidia is taking this threat. Now I'm not going to walk you through all the data here, because this is not a tech channel, this is an investing channel. But just one quote, and this comes from somebody at semi-analysis. They run a great website on this stuff if you want to like really go geeky on this. And they say Google silicon supremacy among hikertha scalars is unmatched with their TPU seventh generation, which is arguably on par with NVIDIA Blackwell, which is the new chip. This is how good this is. So don't write this off. Don't stick your head in the sand. It's also not over for NVIDIA. I'm not saying that. But what I'm saying is there are some opportunities here. We also have Amazon developing two types of custom chips. I'm not going to run you through the whole thing, but similar story, right? Meta has been developing its own inference accelerator. Meta is also in talks with Google to use their TPUs. So you're seeing these guys are banding up together and they're going, why are we paying four times more than we should for these chips? And here's that deal. By the way, Microsoft is also developing a custom chips, but again, I don't want to bore you to detail with like tech stuff here. The bombshell that just came out is that Meta is making a deal to rent Google's TPUs in 2026 and purchase them in 2027. This is one of Nvidia's largest customers, actually, two of the largest customers, planning to use a different chip. It's like McDonald's, Meta and Google, they are head-to-head competitors. Yes, they're both advertising businesses. Anyone who advertises online puts money into Meta and some with Google and they decide which way to split it. They are direct competitors. So it's like McDonald's saying we're gonna buy burgers from Burger King, right? Or the whatever, the bread or something, the bun. So it validates that these Google TPUs have in is that are a legitimate alternative and it shows you the market is fracturing up. Now, when this came out, let me show you what happened to NVIDIA. NVIDIA stock tanked, right? Here it is, Tuesday 25th. There it is. I remember getting that news alert, of course, on my phone. Stock went down, right? It's recovered a little bit since. And if you're fair to it, it's basically exactly where it was in July. So it's really just traded sideways, right? But we can do one better than that. In Trade Vision, you can look at dark pool data. So you can see all the popular ups and downs that Wall Street is trading, but you can just click on trades up here. You can type in NVIDIA here, and you can see what's going on there. And actually, Wall Street right now is buying this dip. Well, they took some profits on some of their puts, and now they bought yesterday$41 million of calls of call options, a lot more than they bought puts, which are the bearish instrument, right? You can see the biggest trades out there, right? The biggest trades out there are actually, well, I don't know what the summer is here, but there is some bullishness there, right? So NVIDIA isn't over, but the story is changing. You can do the same for Google, by the way. You can look that up. Go to tradevision.io. Uh, there's a link down below to it. Take advantage of. There's a free trial. So you can literally test it out for free. Uh, and then if you want to stick around, you you get to take advantage of the crazy Black Friday offer that we've got going on here. But you've got to understand these big tech companies have a common motivation: cost reduction. The 75% gross margin, which basically something that costs$25 to make, they're paying$100 for it. That's not nice, right? You don't like that. Say, you know, you went out, you bought this phone for$1,000, and you found out it costs$250 to make it. You think, well, couldn't I just buy for$250 somewhere else? And you've got all the money in the world so you could actually build your own. They also don't want to be dependent on one supplier. You can just raise prices. This is what they are at the moment. And then third, they can actually build hardware that works exactly for their needs. Because an NVIDIA chip is always going to work for everything. Everybody. It's going to make everybody happy, right? And so this is not a fad. Tesla came out, then if we have it in here, yeah, we do. They just announced, Elon just announced, that they're going to produce more chips than any other company starting in 2027. Seems far away, doesn't it? Why do we care? Because it creates a new competitor. And the stock market, generally speaking, is looks forward. Maybe not two years, if I could spell forward, but maybe one year. So at some point in 2026, this brick is gonna hit people in the head. But you already know it. You've already set up your alerts on TradeVision, you already are informed, you're ahead of the curve, right? And that's my goal here for you. And maybe Tesla will come out and go, well, you know what, we're gonna sell these chips, right? So everybody who wants them could be a whole different business for Tesla that could compete with Nvidia. Google looks like they're already doing it, right? Microsoft is gonna have their own chips and Amazon and everybody else, and maybe they all start selling them. Then you suddenly have not one massive NVIDIA and a struggling AMD and a basically inter-on life support. You suddenly have Microsoft, Meta, Google, Tesla all become chip companies, all selling them to everybody else. Do you still think that Nvidia can charge a 4x markup on their product at that point? You're gonna think that through. So for me, this is what I'm seeing. I think 2025 is still gonna be great for NVIDIA. I think 2026 will still be insanely great for NVIDIA, but we're gonna see some changes. So why am I saying that? Is this not a doom and gloom video? Is it not all over? Look, making custom chips at scale takes time. These companies are in the early stages, right? Nvidia is gonna have a massive year in 2026. But I think that is the inflection point. I think that's where we're gonna see that scale is gonna come from the new competitors. You're gonna see Amazon's training three in the market. You're gonna see Microsoft's Maya 200 finally deployed, unless they cock it up again. Apparently, you're not meant to use the word cock in the same sentence as Microsoft and Bill Gates. I don't know why. Now the Google Meta deal could be beginning. Meta is actually then using TPUs from Google, and they're gonna talk about buying them and tell you when. These are big, big shifts. And then 2027, the real competition materializes. Meta starts purchasing these Google TPUs, which means Google has enough to sell them, right? Tesla's chip production scales up towards their crazy goal. So when's the market gonna react? The market generally is six to twelve months forward looking. The market isn't that smart. So the market, in my humble opinion, and that's all this is, starts pricing this in sometime in 2026. If there are delays, because it's difficult to build new chips, it could be 2027. But currently we have pretty lofty valuations for these kind of stocks. So I go into better stocks GPT, which is part of our free community, um, at least some limited access in there. I see a PE ratio of like 56 for NVIDIA. Yeah, so I just typed in literally NVIDIA. Um, that's pretty lofty stuff. So maybe it's a little lower right now. Well, you know, we're somewhere in that space. What if that PE multiple drops to 35? Now bear in mind NVIDIA is still growing, so it doesn't mean the share price is going to collapse, but we might see a flat stock because we see margin compressions, as in they still sell more, but they're making a little bit less per unit, and therefore the growth doesn't really deliver that much. So there is still a ball case out there. Nvidia is the market leader. They have this crazy software mode called CUDA. Every developer in the world in AI works on CUDA. So you come in with a new chip and you're like, yeah, I don't really want to learn that. That's gonna protect them, right? And they are incredibly good at innovating and delivering, and they do have more diverse customs. But think about it, the top two, 40%. That's the real risk. And then you have the three and the four and the five who are all gonna be part of this party. So as AI becomes less about, I just need more and more and more compute at whatever cost, and these companies start looking at, well, we're actually gonna have to justify our spend, we're gonna have to make some money on this. A cost-effective solution is what they want. The question is, can NVIDIA deliver that? Or is NVIDIA always gonna be the Maserati? But really, most of the time they want to drive uh, you know, see what I'm saying? So, what should you do? Well, look at your portfolio concentration. Are you massively exposed to NVIDIA? If you are, think about doing something about that. And you might want to diversify into some other sectors that are not all AI chip dependent. And I don't mean taking profits as in selling stuff because it's up a certain amount, but stops, for example, can be a smart way of managing that, right? So you're gonna want to stay on top of what the hyperscalers are saying, the Microsoft's, the Googles, the Metas. And again, Trade Vision will keep you informed of that. What you've got to do is select those stocks, uh watch the margins for NVIDIA. That's gonna be the real thing Wall Street's gonna be watching. But how do we make money out of this? There's always a way to make money out of this, right? Well, do we just buy Google stock? Look, yeah, you could buy Google stock. But I would prefer to invest in the picks and the shovels behind the Google TPU, right? So I'm gonna give you some of these tickets here, um, write them down and take a screenshot. I'm not gonna run you through in super detail here, otherwise, this video would get just crazily long. But ABGO. So these are all the companies that are the key suppliers to Google to allow Google to make those chips. ABGO, which is Broadcom. APH, which you've probably never heard of, right? Amphenol. Then the foundries, TSM and AMKR, write these down, write these down. And then you have, sorry, the slightly messy slide, you have the guys who do the assembly, CLS, JBL, and then the integration again, JBL and Flex. And on the networking side, you've got Lite, CLS, and again more ABG. These are your key suppliers for assembly, integration, networking. So Nvidia is an enormously marvelous, incredible, great company. But the unchallenged dominance era, where they're the only ones out there. That's going. It's a bit like imagine you invented the motor car, right? And you were the only one selling a motor. I don't know whoever that was early on, you know, Ford in the US or, you know, Mercedes-Benz, where I held from or something like that. And you were the only one. You could sell them for whatever price, whatever you built, you could pre-sell, it was as glorious. And then you fast forward a little bit and competitors start sprouting up. And that's what happens when you have crazy margins, right? So those are the risks. All I'd say to you is stay informed on this one. Take advantage of our Black Friday offer at feedexfriends.org slash tradevision. I literally think it's like, is it something crazy, like 11 bucks or something? I don't know. Normally it's like 19 bucks. That's really super cheap anyway, but it's it's even cheaper. Um, and it'll help you be informed of these things before everybody else finds out. If you got some value out of this, Sharab is a friend. And I wish you great success.