The AI Trade That Could Explode After Nvidia's $500B Bet

Coreweave just reported earnings and the market is pretty happy. Neoclouds are popping again with 20% candles across the board today. But how do these companies actually work? And how long is it going to take for them to actually become profitable considering that they are spending an ungodly amount while also making a lot too? And why is one of the most famous economists in the world shorting them? What's up everybody? It's LG Ducat here and welcome to Milk AI, the daily AI show that recently moved to 5 days a week and now I'm overwhelmed with information. Today is August 12th, 2026. Recording on the 11th, just an hour or two before core we've reported their earnings. Last month, our episode about Neoclouds was one of our biggest hits. You guys told us that not only did it make you bullish, it also helped you understand what they're all about. And yesterday we ran it back as Martin, our lead researcher, took a closer look at Cororeweave and Nebius and even developed his own model for how to value these companies. Today we'll look at that model, how many years it'll take for them to be profitable, and why it won't take much for them to keep moving up. And while you're listening, you'll probably be tempted to get a look at Martin's Milk Road Pro portfolio, which holds both companies and a dozen more. If you're not a member yet, getting in costs just a dollar for a 7-day trial at the link below. And a reminder that our podcast today is free and that it wouldn't be possible without our partners at Saber.money, the stablecoin payments platform built for Asia. Keep an ear out for more information about them later in the show. Martin, welcome back to the show, man. I'm keen to hear about Neocloud. So, take it away. >> Hey, uh, thanks for having me. Um yeah, I think everyone saw the announcement of Nvidia yesterday and it was pretty much saying that AI compute has now became an asset class and you know Nvidia, Black Rockck and all other big finance financial houses has agreed to finance this whole next consortium worth $500 billion and it's just a bullish news for No cloud and so nobody knows today how to value these GPU fleets but when you have got this Nvidia's announcement that they are going to pick you up meaning that GPUs will have value even after their expected 6 years lifetime that's giving you a signal that no clouds are going to be in a strong position from now on and I want to capitalize on that trend And actually today I want to decide if nail clouds are still a good buy because yes I might be bullish on their business but are they trading at a good price or not? So that's what I'm going to try to decide today and I'm going to walk you through my model and then I'll show you the different methods how you can look at nail clouds businesses to decide whether it's a worth what's what's the price today or maybe it's too expensive. I don't know. Well, I do actually know, but I want to show you how I'm thinking about all this. >> Excellent. I think that that's a that's a question everybody has right now, too, right? Is especially with that announcement, but even before as Neil's kind of waiver, come back off their all-time highs, what how do you value them? So, I'm glad you put something together for that. >> All right. Um, so I guess everyone knows what new clouds are. I'll just, you know, um, simplify that. It's those are pretty much just public companies who are buying a lot of AI, a lot of GPUs. They put them into data centers and then they go out and they rent those that computation to hyperscalers to you know other businesses whoever and that's their whole business. And here on the very first slide you can see some numbers behind that business because on the first side it might not look that good because as you all know this year so far has been punishing hyperscalers not because people are bullish on their business but because they are burning a lot of money and nail clouds are no different. They are going to burn a lot of money and I actually think that they are not going to be profitable until 2030 or beyond. So, you know, that's one thing I want to say here. Here you can see the table with Corv and Nibbius. I think those are two most advanced nail clouds out there. Their backlogs are huge. Everyone knows them. And if you say word nail cloud, everyone thinks about either corv or nebus. There are bunch of others, but they are much smaller. And you know, I might also want to say that Nvidia has stake in both of these two nail clouds. And if Jensen was signaling something all the time, it was that he doesn't want only couple u hyperscalers to succeed. He wants a lot of companies to succeed and therefore you know he he got some stake in these companies and now they are making this huge deal that should help all these nail clouds and similar companies to really find the money to get more GPUs in use and sell them to the market. So that's the whole idea and all the numbers you can see here are I would say reasonable or maybe even slightly more conservative because those numbers are consensus numbers that Wall Street is expecting and this is pre-announcement what just happened with Nvidia. Okay. So I think we are going to see a lot of you know rerating and a lot of targets being increased for all these nail clouds because of that announcement again is very supportive for businesses like this and so those are the numbers that people use today and I'm going to use them today as well. So you can think of it as this is the base case. This is before that announcement and while we haven't seen any changes from Wall Street yet, I think those changes are coming and they will probably most likely just increase all these numbers. You can see here what you can see here is I just show couple metrics for Corv and Nibbis. You can see that there are 20 year 2024 2025 those are actual numbers. Then there is year 2026 that's the actual year and those are estimates or actually the guidelines from from those companies and then 2027 and beyond those are the estimates. Okay. So that's what Wall Street expects from these companies and you can always like those estimates change over time. So it's not if you build model like this it's never a one time off thing. You always need to update these numbers because you know the market changes, the sentiment changes, the environment changes, interest rate changes and there are a lot of dynamics and inputs that affect everything that we are talking here. So this is the current state. This is today again because of that announcement I guess this is going to change a lot over the next few months to the positive side. So whatever I say today, you can think of it as a very conservative take on new clouds. >> Just going to pause there for a second to point out that the market is showing signs of something kind of different happening and our analysts at Milkro Pro are all over it. They spent the last couple weeks making a lot of trades, getting out of some positions and getting into a lot of new ones, getting ready for the next wave of robotics space or even kind of picking some different AI winners. If you want to see what they have in their portfolios, what positions they're opening, it's just a dollar in Milkroad Pro at the link below. >> Right. And your conservative take is basically that these they don't come into net income because they're spending so much money for 5 years basically, right? And that's that's a chart we're looking at now that they are making a lot of money, right? And even here you have Cororeweave making are these are these numbers in billions? Are these billions or trillions or what is this? >> Yes, billions. >> These are billions. Yeah. um that Cororeweave is like is making $60 billion annualized in 2030. Uh but still that but still spending like an obscene amount, right? So that their net income is still is still in the negative and so is their cash flow. Why? Here's a question before we continue, Martin, maybe for people who are still learning about NeoClouds and kind of the business there. And you talked about this a little bit. Why do they need to keep spending so much money? Why what is it that they spend money on? because from what I understand they already have something that their their customers can use right >> they are sort of run running their business so they need to put a lot of money first so they buy all these GPUs all the servers all their networking all those kind of stuff and then revenues will come eventually but that you know they need to frontr all these cost first before they see all these revenues coming in so it's very very um capex intensive business and um but yeah, it's it's probably going to be worth it to be fair. >> Okay, >> but >> okay, >> now here is the question. If I had a food truck and I want to sell my food truck with with hot dogs and I'm going to say I want to sell it for $100 million, are you going to buy it? Well, maybe you'll say, "Okay, I love your hot dogs. They're awesome, but why should I pay $100 million for that food truck when it only makes $1 million per year? And that's fair. So, you are you are pass you're going to pass that offer. You are not going to buy my food truck. The same thing I tend to do with all the companies in my portfolio, but not everyone does that. So I am bullish on new clouds but I want to make sure that I'm not buying a business that should be worth 10 billion today at the current price that is you know 20 billion uh 50 billion. So that's the whole exercise that I'm going to try to do here with all of you and and share my way of thinking about it. >> Okay. Okay. Great. Well, let's talk about your model that you built here for for basically how to value them based on that analogy you just gave us about hot dogs. It's a very relatable example. >> Well, by the way, I wanted to have food truck when I was younger with my best friend, but um didn't work out. >> It's not too late, man. You buy buy enough nebas and if it does what you're saying, then you could open any business you want. >> Well, you don't know yet. Maybe it's it's too expensive right now and I'm not going to buy Nebius. >> Okay. Anyway, um so I have three different valuation methods that I I'm going to use here and explain with all of you because you cannot apply you know the common price earnings ratio because they are not making um any earnings like the earnings are negative so you wouldn't get any multiple you cannot use all the traditional you know revenue multiples or um aida multiples because those companies are copex heavy and so they are not going to make any profits or these valuation methods do not include those copex and so those are not good methods to value companies like this. It might fit for banks or maybe software companies but not for these ones. So you really need to think about okay how to value these kind of companies. I have three methods. First of first one is because I'm investor I want to make sure that they make attractive returns on that equity capital that they got and so they are probably going to do a delusion uh like all of them because again you might need a lot of money here in this example corv they need 30 million equity for one megawatt but they can only get that 75% of that. So the rest they need to pay for it. They need to put down the cash. So usually what those companies do is they are going to issue new new shares and they are going to sell them in the market. So they are pretty much deluding me but I want to see that okay I'm fine with that if you show me the returns are good enough to do it. And so here I just uh show to here two tables. One is for Corv, one is for Nibbis. And there are two assumptions you need to make to really figure out if there are going to be, you know, good returns on that investment for you as an investor, if you are going to get diluted or not. And those two kind of key inputs are what's the GPU lifetime because today when you buy GPUs, everyone is depreciating GPUs in six years. So imagine you buy GPU, it cost let's say $1 million and over those six years that GPU is going to have accounting value zero. Okay. So now what what is the expected lifetime here? Actually with this new announcement of Nvidia, Jensen also argued that even 2020 chips from Nvidia are still valuable and useful today. in 2026. And so his argument is that even older even chips beyond that six years lifetime can still be valuable on the market. And this was a huge message for all the lenders because if you think about the first six of those contracts and you know that they are going to have some value even after that and actually Nvidia is guaranteeing that 25% residual value will remain for those GPUs. That's pretty strong message. So I would be okay to earn them money because you know it's backed by Nvidia 25% it's not you know 50 60% but it's good enough. It's it's sort of a your downside is limited. So here my best case is that I think the GPU's lifetime could be something about 10 years and I expect that the residual value could be 60% after that you know six years lifetime. And if that's the case, you get you will get 20% internal uh return um on that equity. So for me as an investor, that's awesome. That's what I want to see. And I would say okay guys, if you are going to keep using these economics and these terms for like your business for additional megawatt you will bring in, I'm happy to do it. So just go ahead, you know, get more depth, build build more data centers, buy more GPUs, and you can continue with that business because this makes sense. With NIBUS though, it's a bit different because I should also say that Corv is leasing all these data centers and these shells. they don't they don't own those buildings but Nibbius is is having sort of different approach and they they own 80% of the buildings that they are using. So that's a little bit different story. So obviously copex are going to be much higher because you don't only buy GPUs but you also need land and you know building and do all that stuff. So your debt is going to be much higher and also you might need to raise more equity which makes that you know returns less attractive than with Corv. So here your goal is if I believe that those two companies okay maybe I should also say what's my best case for Corv uh for NIBUS sorry it's actually the same I think GPUs will um live like have value for 10 years and at 60% return is 12%. Is it's still decent but is it enough to like you know risk my money? It's it's less obvious I would say. So my goal here is okay if I believe believe in these assumptions I think Corif is a better choice here and I just need to make sure that the contracts the the costs and the revenues that they are making from these megawatts will will hold and if that's true that's fine I can see the return for me as an investor. So that's good here. I would pick Corv. Do you have any questions here, LG? >> I was just I was just going to summarize that basically I hadn't thought about this of of how much of like specifically the buildings that they own, right? And how much they have to finance that and that that is a I think a lot of people saw that as a positive for Nebius, but you're painting as as a negative as well on the balance sheet, right? that it's like that that is that just requires so much more like they're that's more debt that they're taking on for that and that that is a deterrent and that that is is is worse for them versus core just straight up renting. >> Yes. But at the same time you could argue that what if Corv, you know, Corv rented from Galaxy War 15 years, what if they after the the GPUs will wear off in in 10 years, what about them? they still have 5 years gap. So they will need to find another GPUs that they will put there because they are paying that rent anyway. So you know it's it's a good tradeoff I think but you cannot probably say this one is better than the other one. It's just a different approach. We don't know yet which one is better. >> That makes sense. My other question for you about this slide is that I also hadn't really thought about the GPU life being such a huge part of I guess the question around these businesses, right? Is that it's like well we don't know it's hard to say how long those GPUs actually last from here uh and when they'll have to get new ones and have to replenish that, right? And I guess that's kind of like that's what is that why Michael Bur has a short basically like on these companies is because he's like listen the GPUs are depreciating way too fast. these companies are going to be totally screwed because they'll just have to get entirely new GPUs like like in the next year or two basically like he thinks this is a really short-term problem for them. >> Yeah, he thinks actually that GPU's lifetime is something about three or four years and you can see here that >> if if you just look at Ky for example 6 year and 0% residual rate you will get at 11%. which is not that good, right? So like >> but I think that that guy who is like you know shorting a lot of AI stocks he's not doing well I think and now he was certainly not happy about that announcement or it was actually announcement >> because it just killed his whole thesis >> right yeah of course yeah and and and I mean he's mainly known for one big >> short that he figured out other people didn't you know so this is these are these are wellknown even what you're sharing it's like this isn't some new science, you know what I mean? Like it's not some secret that he's only done the math on. Like this is this is well known and and people know these risks and are still choosing to participate in the market and buy at these prices anyways, right? So um you know anyways. Okay. >> Okay. This makes sense. So this is a good this is a good way to value them and you're kind of surfacing a lot of the the the factors that go into how you can actually look at these their balance sheets for both both of these companies. >> Yes. And then there is a second method which is a little bit similar but it doesn't only look at my um return as an investor but it's looking for a business returns. So it's not only looking at equity that investors need to put in but it ignores whether it's equity from shareholders or if it's debt. So think of it as every business needs to make some value and if I think that these companies are going to make more value than let's say what average cost of capital are then that's great I want them to do it but if not well maybe it's not a good business at all and so maybe you know I shouldn't buy these companies in the first place. So again here I have corre and I have nibbis and what we can see here on the y-axis we have depreciation rate which is something similar. So you know if depreciation rate is 10 years we are pretty much saying that the GPU's lifetime is 10 years that's that's the bull case that's what we all want. Well, definitely that's what all these new clouds want because the longer the better, but you know, right now everyone um accounts for 6 years lifetime. So, it's something around 16% depreciation rate because if you divide, you know, 100 by six, you'll get something around 16%. And then if if we look at how much money they are going to make in our base case in 2032, we can sort of calculate how much money how much capital returns that's going to bring over the years. And so here you know on the very first slide I showed some numbers and here if I just put them into this formula and again I assume 6 years lifetime for these GPUs as a base case I I might be a little bit more bullish as I believe that they will have uh 10 years lifetime. Um but anyways the point here is the base case shows me here that for that whole business what Kore is doing it's actually making a good return of money because even after you know I deduct all the all the cost of the money I think I used 10% um that cost of capital I will get still positive returns for the whole company as a business I'm not only talking here about that equity because again you know you might the difference between those two methods is that in the first one you might get you can be tricked because the company might use a lot of leverage and so your equity returns might might look very positive and you might be like wow this is awesome if they are making 30% you know returns for me as investor that's great I'm just going to all in But then if you run this method, you might figure out, oh, hold on. But the returns were really good, but because they were using a lot of leverage and the whole business is not actually making any value because the returns that they generate don't even cover the cost of capital. But it's not happening here. So the base case here is pretty much saying that okay this business is they also use leverage but a decent leverage and they they can just make returns that cover cost of capital and something on top. So that's also good here. So we got you know the first method was saying okay that's pretty good. It actually made me more bullish on Corv than NIB is. Um, and this second one is also um showing that Corv looks better. And you can see that return the base case return on invested capital is 7.5% for Corv versus just 2% for NBIS. So it's a it's a again Corv is a bit more attractive than than Nibbus. You know, one thing we've talked about a lot on this show is that crypto is quickly becoming a huge part of the global payments infrastructure. And nowhere is that more obvious than in Asia. But if you're actually running a remittance company or a payment business, you know that the hard part isn't moving the stable coins. It's dealing with local banking partners, compliance, liquidity, and all of the operational headaches that come with sending money into places like India and Southeast Asia. That's why today's partner is Saber. They give payment companies stable coinpowered infrastructure to collect and make payouts across Asia without having to build all of that complexity themselves. They've already processed more than $3 billion in transactions across 40 different countries. So, this isn't just a concept. These guys are actually doing this for real. If you're building payment infrastructure or expanding into Asia, make sure you check out Saber.money. >> Is the market reflecting this right now, Martin? Like this this this type of calculation? No, it's not. But is it is it would you but but would this just make Nebius less bullish but less bullish than coreweave or would you make it bearish on Nebius and bullish on Cororeweave? I think I am personally a little bit more bullish on Corv and the reason is that I think people don't like Corv because they think that they have too much leverage at this point. they see you know 30 billion debt and people are scared but I will just say that Nibius is two years behind Corv meaning that in 2028 I think Nibbius debt levels will be similar to today's corve levels so you know if people really prefer Nibbius today because they might still have I think they have 9 billion in cash and they have also So something around 10 billion in debt. So they are pretty much net net you know um around zero. So that's what people like today but in two three in two three years it's going to be very very different. I'm pretty sure that the depth levels are going to be very similar and so people really need to know that again core was the very first new cloud in the business. They pretty much just create this whole market and the whole structure for this and Nibbius is now taking advantage that it's already built. They still have a cash that's fine but you know at some point the the financial metrics and the debt ratio metrics will be similar to um curve today. what what out of all uh these are fantastic numbers and I think this is this is a really good way to frame it too that like longterm it's it's it might be a little less relevant but what in your opinion I was asking you kind of how does a market view this what narratives does a market actually like latch on to right now because in the last slide we were talking about you know obviously someone like Michael Bur is really famous so people take that short with some level of authority but that's not the only thing that's being said like are are is the market seeing it this way or is the market just purely reacting to all-time highs and and like famous shorters. >> I think so. I I I don't see any rational like I think it's just market doing its own thing where it's more about the story and price action rather than you know give me the reason why Nibbius is so much better business than carviv and you might struggle to do it. So again I think it's more about Nubius might sound more sexy better storytelling guys you know better marketing and so but I am believer that over the long term fundamentals will be reflected in prices and so all the short-term or mid-term volatility we see today around these snail clouds is going to be you know less important and unless the fundamentals still hold and the the business they are doing still holds then I think corre is actually having much more upside than Nibbius from from these levels. >> Mhm. Yeah. And like you're saying I guess you're also alluding to the fact that they're they're going to be bundled together as well as like the leading NeoCloud companies. So that this this level of detail >> is important when you're choosing between them, but that inevitably like they they will move together as Neoclouds as like a category. >> Yes. >> Right. >> And in fact, you could argue before that announcement that this is pretty much I I think about Neocloud th those are leveraged bets on hyperscalers. They do similar business but they only focus on GPUs. But if you really wanted that leverage and you really think that GPUs might have extended lifetime after the six years then I think corv is is more clear bad because they are not buying those buildings. all the capex is just you know buying GPUs and so if that's your trade you should maybe you know think about corv more because they are not buying land they are not owning any data centers and they just own GPS >> right yeah okay yeah that makes sense let's let's look at their actual valuations right now man so and this is something where core is currently valued at basically twice in nebas That's a current that's the current that's their current market cap. So I mean it's not they're not apples to apples. >> But here I have asking price. I don't have market cap. >> Oh. Okay. My bad. You go ahead. >> I actually think market cap is quite similar today. But if you buy business that has a lot of debt, you are buying that debt with that business. So core beef might have market cap 50 billion but then you need to add that debt levels that they have. So let's say 27 billion and that's the enterprise value that you are buying you know. So you might see something else on market value but it doesn't include that debt. And if we are talking about debt intense companies you always need to look enterprise value. That's the value of the company you are really buying not the not the market cap. That's why you see that some people use um revenue divided by EV which is enterprise value or you know um AIDA divided by EV enterprise value because market cap doesn't make sense here and so here this third method is reverse uh discounted cash flow and and the reason why it makes sense here is because those companies as I said earlier are burning a lot of money, a lot of cash. They are not making any profits. So you cannot apply, you know, discounted cash flow in a in a way that a lot of people do. I I'll do it too, but here I cannot apply to the company that's not making any money. But what I can do is that I can sort of take today's price um and then project it to the future to see what the what will get me there. And it's sort of a reverse engineering. So it's not that I'm trying to estimate the final price. I'm just trying to estimate what the market assumptions are today. So what growth is sort of expected here right? And so when I use this um method here you can see that for corv again as I said asking price is enterprise value my base case is 10% weight average cost of capital you again need to assume that there is some cost of capital and then you also need to take into account okay so it's 2032 and what do you think that I will be able to you know exit that position in 2032 and at that point it might already be mature business they might already be making a lot of money they might be making you know positive cash flow so you might really use here exit multiple of their AIDA earnings so I think and that it's reasonable to assume that in 2032 the I might be able to exit this company at eight times AIDA which is pretty decent. It's it's reasonable. It's not like I'm not trying to create any magic. That's how these companies are. Once they are matured and they have some positive numbers as well, they are not just burning money. You can use these multiples for for your exit strategy. So I think my base case is that if we assume that 10% is the blended cost of capital and I think I'm going to exit at um 8x of their AIDA in 2032. It might require uh revenue at that point to be at 51 billion. I'm looking at corv 8x on the x-axis and 10% on yaxis. I will get that multiple 61 and it tells me okay my base plan my on that first slide I showed you that I think the revenue for curve is going to be 84 billion in 2032 and if I use these assumptions here it's actually projecting that I might need just 51 billion revenue needed did. So it tells me okay so this is actually not expensive because um the numbers that I can reverse engineer into today's value are actually quite attractive. So it's pretty much telling you when it's around one multiple that it's it's fair valuation. But if it's below your base case is telling you that okay this is pretty interesting. And pretty much all the green all the green cards here is telling you that okay this might be an attractive price. So it you can still buy it. it it doesn't tell you like it's cheap you know you might think of like I might deduct one minus this 61 number or whatever no it just tells you these companies are actually have a reasonable you know forecast for the next 6 years and in our base case it's not even needed that much revenue making in 2032 to make it worth it today. That's what these tables are telling me. So, I can see here again my base case for Corv is telling me, okay, this is good business. I should buy it. So, and again like surprisingly corv is again you can see a a lot more green cards on Corv than on this. So, um this this is it. Um pretty much corre one on all three methods and it tells me that they are making a lot of money for you know investors from a business perfect uh perspective they are also generating value to the whole business they are able to beat the cost that they are paying to run that business that's great and also now I don't know what the future value should be but if I reverse it into the current uh current present value. It's okay. It's also a good price because I don't even need that much revenue in 2032 to make money out of this trade if I buy today. >> Right. So basically, so let me just wrap my head around this. So basically what you're telling me is that even though right now the this the price of these may suggest like some level of uncertainty or weakness like from the market they to to reach like a much higher valuation. They they their revenue numbers are like if they can even get close to those revenue numbers that you are suggesting then these will repric like pretty aggressively right. Right. Is that is that basically what you're telling me that it's like if they can even get close to like that 60 billion like you were saying for coreweave in 2032 then it doesn't it doesn't take a lot of that to even happen for these to actually move up a lot. And that's basically the opportunity you're outline outlining right that it's like a lot of these revenues are projected. It should hit it so long as you believe in the rest of the AI trade and that all this stuff is going to continue to move up and you like Nvidia's announcement and all that kind of stuff. Um and yet and yet even a fraction of that is not yet reflected in the price of the companies. Yes. >> Cool. Okay. Okay. That's a good that's a good way to look at it. Well, why not though? I still don't understand why not. Like why does the market not want these though? Like does it like versus other sectors of the like the bottleneck stocks or the buildout? I actually think that we are at the point where the sentiment is changing and I think people are now starting to look at hyperscalers again and you know some of them I think Microsoft did really well uh lately and I actually bought um I want to buy more Google as well. So I think we are at the point where people are going to realize that capex are only bad if you look at the burn and yes those companies are still going to burn a lot of money but you should not be worried about that if you think that there is demand for it and that there is demand that there is return for that capital and for hyperscalers there is it's like their backlogs The total backlogs are like 500 billion. That's just insane number for hyperscaler and there are many of them. I think like corv has a backlog of 100 billion. I think nibbio is a bit lower but I think they're going to get also you know backlogs north of 100 billions. So this is very interesting today and I think that the more bullish people get on um hyperscalers and now with the tailwind we got from Jensen that Nvidia is pretty much telling you that the the lifetime for GPUs is going to be you know beyond 6 years. This is the moment you want to look at Neoclouds because they give you that exposure to that market. And if we started the conversation today asking that the key question for nail clouds is GPU's fleet and what's the value of that fleet after 6 years. show the the lifetime. Well, if we if we believe that there is Jensen Huang Nvidia CEO probably the smartest guy in AI world and like not like the whole world but certainly he knows something about AI chips and what they residual value should be. So they are not just going to be wear off in six years and that's it. You can just throw them out. that's not the case. So I think that's that's the moment for you know all these new clouds to really start popping and I think as I was trying to highlight today the prices still make sense and so I think I'm going to add more actually. >> Oh wow. Okay. Yeah. So this these are two these are two companies that you already hold in your Milk Road Pro portfolio and the move from here is to keep adding them. And I like what you're saying too. So like, you know, inevitably you have to be listening to Jensen who naturally has an has an interest in these companies continuing to move up, but he has said like specifically Neoclads and I think he does he does he own both or does he own some Nebas? >> Both. Yes, >> both. Exactly. Yeah. So these are parts these are companies that he sees as like quite vital um and probably seeing some of the you know even some of the projections that you're you're looking at right now. >> Um that's great Martin. Go ahead. One more thing I just want to say. It wasn't just that Jensen just gave me sort of a green light that GPUs are going to be, you know, have extended lifetime, but it's also that the costs for these nail clouds are going to be much much lower because today they are, you know, borrowing against GPUs and, you know, some lenders might be okay, but what's their lifetime? But now, hey guys, Jensen just said that they are going to have extended lifetime and if not, Nvidia is backing them by 25% on some deals. It's not like a general rule, but you know that that message that signal is super strong and so I think people should really pay attention to it and like I couldn't think about any better signal for NeoClouds than this one. >> Okay. Well, that's well said, man. And I think that that's good. Um I think for me as well, you know, I've been trying to understand and I asked the guys yesterday on the rollup too, like how does this $500 billion manifest? Like, you know, how does that how does that actually come into the market and and where should we look? And and naturally, Neil clouds are clearly a spot for that. And um it's great to see this kind of model. I feel like we don't really do this enough on the channel, right? Like we talk more news and some of the more standard fundamentals, but it's great to see like a higher level of math put into it. Uh so I really appreciate that, Martin. And if you guys want to chat with Martin or see his portfolio, all that is in Milkroad Pro, so make sure you check that out at the link below. Um, otherwise Martin, uh, we'll see you back here next week. >> Yeah, it was fun. Thank you, LG. Have a good one. >> Want to stay ahead of the biggest technological shift in history? Subscribe now to get insights straight from the sharpest minds in tech and finance. Quickly, you'll note this show is for educational purposes only. Nothing here is financial advice. Investing always carries risk. Never invest more than you can afford to lose. Thanks for tuning in. See you in the next one.