If You Are an Nvidia Shareholder...Get READY!!

Welcome back to the channel. I'm Jose Nahar of What the Chip Happen, the greatest, in my opinion, semiconductor and AI YouTube channel online right now. Now, today I want to take a closer look at my favorite semiconductor company. You guessed it, Nvidia. Recently at the SpaceX earnings, Elon Musk shared some amazing amazing information that Nvidia investors should know. Now, before we get there, we also do have some important things that Meta and some other AI companies like DeepSeek just announced that you don't want to miss. So, let's take a closer look in today's episode. But before we begin, if you're serious about semiconductor and AI investing, I break down major earnings and conferences inside my community. Institutional quality research built for retail investors. Get 33% off at whatthechipappen.com. All right, guys. Happy Friday if you are watching this when I release it on Friday morning. The market is doing extremely well, especially for Nvidia. Nvidia now back over $220 pre-market. Super super happy. This is a $5.3 trillion company. I am a huge huge bull of Nvidia. And this is one of those stocks that before the low beta names were the high beta names were drowned or kind of this AI stock crash, Nvidia anywhere near the sub 200s was one of my favorite opportunities in the Mac 7. Now the first thing I want to mention here with um this AI episode is Mark Zuckerberg, the CEO of Facebook released Muse Code and announced is in beta today. It's a terminal coding agent that takes on complete software engineering task across large repos. Now the reason I want to release this is Mark is really stepping up their game at Meta with the AI solutions, right? And if they are really stepping up their AI solutions and AI innovations, what do you think it's going to do for the rest of the market? Right? Companies like Anthropic, companies like OpenAI, even Grock is stepping up. But when competition is out there, it forces all these players to continue and continue to innovate. And that is amazing because with more innovations, we hit new markets. With new markets, you need more and more AI compute. So for me seeing this competition and seeing the story right now even though it is mainly anthropic and open AI you are seeing Grock you are seeing Muse from Meta and they are are are are creating this competitive landscape that in my opinion is going to be extremely extremely bullish for the overall industry. Now there's always always this bearish case that companies like DeepSeek and these lower training models are going to reduce the AI compute because they're so cheap. But Deepseek just recently announced that they plan to raise the overall pricing of Deepseek API servers in the near future with a significant increase expected. Main reason is there's not enough compute. When there's not enough compute, what becomes expensive? The intelligence, right? Mark Zuckerberg at Meta's earnings call actually mentioned it, right? He mentioned that for him, he wants to sell compute. He doesn't he he he he knows he has the opportunity to sell the infrastructure right to sell the data centers and the compute but what he really wants to sell and I maybe mistakenly said something earlier what he really wants to sell is the intelligence because the intelligence is what drives high margin products and we can see here deepseek is increasing their pricing because the demand is so high and the supply is so low and when I'm talking about supply I am talking about AI GPUs about AI compute So this is extremely extremely bullish for the rest of the industry. Now I want to go back into Nvidia now that we saw this whole AI ecosystem in my opinion give us more and more bullish indications that it's going to last a long time. And quick warning just because it lasts a long time. It doesn't mean every day stock prices go up and it doesn't mean that there's not going to be a stretch in valuation on both sides on both the the bullish and bearish side. We are going to continue to see high volatility in this space. But um Elon Musk also mentioned the following right if during the SpaceX earnings call. If you think of where AI was 12 months ago, 12 months ago would have been August of last year, that would have been before claw 3.5, which I think should be considered one of the milestones and a credit to Anthropic. I would say a shock to the system with how good clog 3.5 was. That was 10 months ago basically. So what he's trying to say is AI is innovating so quickly. Most of us completely forgot about cloud 3.5, right? We just started talking about Codex which came out in November and that drove a massive demand in AI compute. So it's pretty much where do you think we're going to be in AI au few months from now, right? We had Codex happening in in November or October. Um we had claw code around the same time before that. Then we also had agents come out with open claw in January or February. We're getting we're already in in kind of June and we're getting kind of agentic tool calling in expanding dramatically what's going to happen in the next few months and all these solutions are going to be additive additive to the amount of compute that we need and this is why I am very very excited that I am investing in this market more importantly investing in companies like Nvidia because I do believe they are going to win. I also invest in a lot of high beta names in this space as well. And if you want more information about the semiconductor industry, make sure to check out my community whatthechipappen.com. This is the community where I do deep dive reports on earnings. We just did a deep dive report on AMD, shared my overall thoughts. We have deep dives on conferences that happen. We have courses. We have a lot of great community members. We have exclusive live streams that no one sees here on YouTube that focus on the AI and semiconductor company. We talk about companies that unfortunately unfortunately won't do that good on YouTube. So, I can't drop them here, right? So, there's so many companies in the semiconductor space that we cover in these live streams. Check it out at whatthechipappen.com. Especially right now that earnings are happening, you want to have that edge before that next earnings call, before that next conference move. So, make sure if you're looking into semiconductor and AI research, that's the platform you take a closer eye on. Now, let's jump into the real topic at hand, right? Nvidia and everything that before I talked about this actually was a subsegment of Nvidia, right? Because all that all those topics, Mark increasing the competitive landscape in AI, Deepseek increasing the pricing of API, Elon talking about how AI continues to evolve at a rapid pace. All that give strong indication that AI compute is going to be needed needed heavily for the next few years. And during the SpaceX call, he did Nvidia something. He he he he did even better for Nvidia. He said SpaceX has committed to using Nvidia GPUs exclusively because they are the best. That is insane, right? That was one of the real bullish reasons the stock was up yesterday and I think it's why the market is loving it because when we look at the numbers it's going to be pretty insane. So at first in the numbers he mentions that by 2020 the end of 2026 right the end of this year SpaceX would have about 2 gawatt of AI compute. He mentions by the end of 2027 he has he expects to be somewhere between 5 to 10 gawatts closer to 10 right he didn't give an actual number but he just said closer to 10 within that range of 5 to 10 assume let's assume he finishes this quarter this year with 2 gawatt he only goes to seven the midpoint right not really the midpoint but close to the midpoint um between that 5 to 10 of next year that's 5 gawatts added Jensen Juan The CEO of Nvidia has told us before that you usually get about $40 billion per gigawatt on revenue and in newer generations is actually going higher. Uh he says it's closer to 60 eventually going to 80 and in the near future it could go into 100 g hundred billion dollars per gawatt. Now let's just do simple math. Let's say 50 gawatts, right? Which is undershooting what uh Jensen has said. If you add 5 gawatt in one year of AI compute and 5 gawatt is $50 billion per gawatt, that's $250 billion per year from one customer from one single customer of SpaceX on just the data center side. Imagine there's not just one AI company. There's Meta, there's Antropic, there's OpenAI, there's Google, there's Amazon, there's Microsoft, there's all these AI startups, there's all these Neocloud players, there's Oracle, there's the enterprise market. If you're telling me the floor with one customer is $250 billion and we have heard Jensen actually say that AI companies right that enterprise is about the big hyperscalers the big neocloud companies there are closer to 50 what is the opportunity for Nvidia in forms of revenue for next year I don't think the market is prepared and they go on to say that hey look because of all our recent acquisitions and the development ments that we're making with Grock, the acquisition with Corser, the demand is going to continue to grow and obviously SpaceX is also kind of playing the Neocloud space um which is extremely exciting as well. Not only that, SpaceX is also doing something pretty cool. They have C they are designing Starmind AI satellite which will be essentially an optimized Vera Rubin and VL72 computer and that is not some sort of thing in the distant future. Now Elon's giving us a timeline. So maybe it's not exactly that timeline, but they expect to start launching this next year. Now there's a cool video here. So SpaceX is partnering up with Nvidia to design Star M Star AI1 satellite compute payload. Each of these satellites will include Nvidia Vera Rubin and Vera CPU Nvidia Rubin GPUs and Vera CPUs for data center class compute. So this is the first step that we're seeing for space AI data centers. Now I'm not that bullish in these data center spaces, but it is cool to see innovation happening, right? It's getting closer and closer regardless terrestrial data centers are still expanding. And Elon Musk actually mentioned that this kind of design that they're working on here for the form factor of these satellites for the Vera Rubin, they're actually going to be using them on their Earth or terrestrial data centers as well. And to me, that's extremely smart, right? Because if even if they don't send this to space, they're still innovating the form factor. And if you're innovating the form factor and somehow increasing efficiency um o of how much power you use, how much cooling you use or whatever um due to the new form factor, you're overall helping the industry as well. Uh so to me this is extremely extremely bullish in the overall space. Now there was something also that I found interesting when SpaceX and Elon Musk mentioned that they were going exclusively with Nvidia. Two big players at Nvidia retweeted it. Sunonny M Madra and Jonathan Ross. For those that are not familiar, these two players are actually from the Gro acquisition, Aquire, however you want to call it. The Gro Jonathan Ross is the CEO and the designer I believe and and Sunny was also a very top level management player. They both kind of retweeted this. Now for me they could obviously just have pride in Nvidia but if they coming from Grock I am curious if SpaceX is going to be a big purchaser of Gro 3 LPX. Now, for those that are not familiar with Gro 3 LPX, this is when you buy Nvidia's Vera Rubin stack solution and you add an extra rack that is going to have Gro 3 um LPUs, right? And this is meant to hit a very very interesting market is extremely low latency with massive throughput. And the way Nvidia breaks it down is here's the Nvidia Hopper, here's the GB 200 and VL72, here's the Vera Rubin NVL72. And when you add when you add the Gro 3 LPX, you're actually able to go further out in this chart. This is where you add kind of the LPX um you're able to go further out through this. And the thing that that opens up is a whole new type of revenue opportunity where you're able to sell ultra ultrabased AI. And this is AI that's super intelligent, super fast, and it's something that people are willing to pay a premium for, right? A premium for an ultra premium for. So for me, SpaceX is going somewhere pretty far and they do believe that they are going to need, if this is true, the LPX system. Uh, so something I found interesting. Let me know if you think my thoughts are a little bit weird here. But with the LPX, it does open up a new opportunity for the AI market. And if these two people retweet it, maybe is a hint that Vera Rubin LPX3 is coming to um, SpaceX. Uh, we also have some other news coming in for SpaceX that I think Nvidia and AI investors should know. Elon during the call mentioned that the memory output is increasing by around 20% per year. Now normally that would be fantastically fast and amazing for any large mature industry because remember the memory space is insanely big and it's insanely mature but asking yourself is the demand of AI increasing by 20%. No, the demand is increasing by 200% a year, maybe even higher. And if you got this demand increasing much faster than supply, economics 101 would suggest that price increases. It does not decrease. And this is not just memory pricing increasing, right? If memory prices is increasing, that means that you have more demand for AI compute. You have more demand for AI intelligence because you're lacking the supply to meet it. And if you have all this demand, what ends up happening? you end up charging more per GPU hour. You end up charging more on API prices. So, it's a very very great time to be a NeoCloud player right now and especially in the near future if this supply is only increasing at a factor of a level that demand is increasing, right? Very bullish for all the AI companies, for anyone building AI infrastructure. the it it's gotten so good that the CFO has mentioned that when with this supply demand imbalance in the compute market they expect it could to continue the payback on some of the new capital deployed becomes less than one year less than one year payback on compute that is extremely extremely wild remember most Neocloud players are paying back in like three years we even kind of heard that from Amazon um certain Neocloud players are a little bit longer than that um some four or five years but SpaceX is saying that we can do it in one year. Now the reason SpaceX is different from other Neocloud players is SpaceX has massive infrastructure fleet. They don't have a lot of small fleets. They just have one giant infrastructure. Well, two now Colossus one and Colossus 2. That brings a premium. Being able to have something that big is a premium in itself and it's an individual market. So, it's not like everybody, it's not to think about it that every NeoCloud player will have a one-year payback time. That's very, very important to remember. Uh, the other thing um that we want to we're seeing is Elon is mentioning that they're trying to get as much power as possible by the end of next year. even though they're thinking about 10 gawatts of active GPU power, they actually want to have about 20 gawatt of actual power for cooling um and for compute availability not so so this means that they are they see power as a limitation they're trying to buy as much as possible even if GPUs they don't have enough GPUs at the moment because GPUs are expensive at least they have the power so as they keep growing in this market they can keep buying and power is not a limiting factor for them. But if you are hoarding up this power, it also forces the other players to really look into more power and make this power issue uh a kind of a bottleneck continue to expand because you keep making that choke point and choke point tighter and tighter. Um and then Elon Musk finished off by saying that he in based on his understanding Nvidia they will receive a large significant percentage of Nvidia GPUs next year. I wonder if if Jensen is getting tired of all these other players trying to go with other companies AMD designing their own AI GPUs and A6 solutions. and video is showcasing. Okay, you w to go and not buy my products. You want to be Microsoft and said that you can slow down at any time if you want. That's fine. I'm gonna sell my products to who I feel comfortable that can buy my products and is going to continue to buy them. So, I'll make sure that this customer that is treating me right, I'll treat them right as well. Right? So, if if you're Microsoft or some other players, you're going to start to worry right now. Right? If if SpaceX gets a lot of the compute, you go, "Oh man, I maybe I have to put a little bit better orders of Nvidia GPUs because what happens if I'm not able to get supply?" Um, and it's something Nvidia can do right now because there is a supply limitation of how many GPUs can be built. And if Grock continues to develop at rapid levels, Elon can easily say it was all thanks to Nvidia GPUs. And that becomes a massive selling point for this player. So, Nvidia continues to be my favorite semiconductor company. It is the one I hold the biggest position in my overall portfolio. And I am super super excited about earnings. Now, I don't care what earnings, what the price does after earnings. I want just the data, right? I'm here as a long-term investor. I'm not here switching answers or what stocks I'm selling and buying, right? It that's not me. I'm very consistent because I truly believe the AI ecosystem is going to last a long time. You don't hear me crying when Neil cloud companies dump. You see me buying. And that's because to me, this is where my thesis lies. And all the data that I have backs my thesis. I could be wrong and I'm not saying I can't be wrong. But right now, my probability of success, I believe, is extremely high. But that doesn't mean I won't get burned if things go south. But everything happening in the industry right now gets me pretty excited for my companies like Nvidia, for uh my companies like AMD, for my core, and for all the other networking and equipment semiconductor stocks that we cover in this channel or in the community whatthechipappen.com. Don't forget to check it out. Take care. Have a good day.