CNBC & Bloomberg On NVIDIA Stock, Micron Stock, SK Hynix Stock, CXMT - NVDA Update

[00:00:00] Speaker 1: NVIDIA wrapping up investments in the AI ecosystem. [00:00:04] Speaker 2: Open models are essential for innovation. It's necessary for sovereignty, company sovereignty. I see a future where the world uses tons of closed models. I encourage everybody, including my company, to use OpenAI and Cloud and Cursor and Cognition and Perplexity. Use everything that you can because it's out of the cloud. [00:00:26] Speaker 1: The company entering talks with OpenAI to help backstop the company's lease of a $500 billion U.S. data center. Sources telling us here at Bloomberg the chipmaker may provide a guarantee of as much as $250 billion. Frank Lee of HSBC has a buy rating on NVIDIA and a $325 price target. Frank joins us now for more. Frank, welcome to the program. You've heard the word that's doing the rounds this morning, literally, circular. Frank, your response to that, how circular does this look? Is that a problem? [00:00:53] Frank Lee: Yeah, I mean, I think this is not the first time we've seen it. We have, but I think this is the first time we've seen the scale, right? I mean, NVIDIA has started to do deals with the NeoClouds earlier. And so this is the first time. It shouldn't be a surprise, but the magnitude and the location. I mean, this is the size in one place. We've seen Broadcom do a similar backstop deal with OpenAI Anthropic as well. So I think it's becoming more and more the norm, especially as people have questions about, you know, the free cash flow and the sustainability of funding the overall AI CapEx. [00:01:24] Speaker 1: Frank, to that point, does that reveal just a little bit of weakness, some vulnerability in the AI ecosystem that NVIDIA even needs to do this? [00:01:32] Frank Lee: Yeah, I think what you're seeing is that you are starting to see questions being asked. So just putting up in CapEx number, like we've seen, that has worked before in the past is not working in this market. Now, what is positive, though, is that if you look at, you know, even though Alphabet has turned free cash flow negative, the leverage and the debt for most of these hyperscalers are still relatively small. So I think you could see collectively the group able to fund somewhere about over a trillion next year and even more in 28, even if it means raising their debt to equity. [00:02:05] Speaker 4: Frank, why is NVIDIA backing OpenAI's effort to secure this compute? Evidently, it was reported that several other big companies were interested in leasing this particular data center compute capacity, including Google, including Amazon. Why OpenAI and NVIDIA? [00:02:23] Frank Lee: Well, I think if you look at, I mean, OpenAI has needed more, I think, help on funding. And you've seen that over the last six to nine months. They've done other deals. And I think, you know, in the case of NVIDIA, you know, they need to make sure that the momentum of OpenAI stays there. I think you're seeing they've already started to commit to hyperscaler, sorry, to neoclouds. But OpenAI is such a big benchmark and such a big representative of the overall AI trade. And I think they need to make sure that, you know, they can keep this going. So I think that's why it makes sense for them to backstop pretty much the, you know, large company like OpenAI. [00:03:00] Speaker 4: Is there any political aspect to this, Frank, considering that this particular data center is being fueled by a Japanese investment in the U.S. It was much heralded. It's on national owned land in the United States, in Ohio, and comes with SoftBank's backing. So it's NVIDIA, SoftBank, the U.S. government. And can you sort of draw a line through this triangle and give us an understanding of what the backstory is? [00:03:27] Frank Lee: Yeah, I think if you look at it, it seems to hit a lot of or at least check a lot of, you know, a lot of benchmarks. You have, you know, politically Japanese sort of related investment now in the U.S. It's being put in Ohio and, you know, I think in the heartland of the U.S. And so I think it hits a lot of these check marks in terms of, you know, where the investment in the data center is collectively, you know, from many different parties and including, you know, more international flavors. So it just shows you, I think the investment in the States is real, right? And it's a very large project. [00:04:00] Speaker 5: Frank, are you surprised that the stock's not up more? [00:04:04] Frank Lee: Well, I think, you know, the thing is, it goes back to the fact that they have to put up this money. So how much does it drive the overall earnings upside from what's already in the, you know, in people's expectations? That's still a big question mark. I think what this does solve is that, okay, funding problems, concerns, that seems to have gone away for the time being. And so I think that at least means stems kind of the concerns that people have about keeping the AI trade alive. But does it, how much does it add to NVIDIA's earnings? I think it's still very unclear right now, like how much incremental is it going to bring to revenues? [00:04:36] Speaker 5: We were just talking about China's largest memory chip maker, CXNT, and the insane IPO it just had. What's your reaction to that? [00:04:45] Frank Lee: Well, I think, you know, because if you think about the memory trade, it's been one of the best spots to be in tech in the last six months. You know, a lot of memory names have gone up almost, you know, 5 to 10x, right? But China hasn't really had a company that has been in memory. And so this is the best way, I think, for investors in China to participate in the memory trade, right? And so I can see why the interest is so strong, because globally people have been able to invest in memory companies all over the world, but we haven't had one in China. Now, the question is going to be, though, is are we near the peak of it? Because as you've seen some of the memory stocks outside of China, they've started seeing a pretty hard correction recently because of these long-term agreements. So what the question is, will we see perhaps this fade at some point in the China market after a strong day like this? [00:05:30] Speaker 6: But particularly chip stocks are now markedly lower. Joining us for more is Bloomberg's Ian King, who is with me in San Francisco Friday night, trying to understand what was a big AI event hosted by the Korean government. Let's just focus on the SK bit. Half a trillion dollars. I had the opportunity to ask Jensen to sort of just explain what it encompasses. You also did some reporting on what it encompasses. What does it encompass? [00:05:54] Speaker 7: Yeah, that's a good question. And the problem is we still don't know. I mean, a while ago, big numbers were a good thing. It was great. They were celebrated. Now, you need to bring the details. So this could be spending by SK on the data centers. That would be SK Telecom. This could be spending by NVIDIA on SK Hynix's memory chips. This could be investment by NVIDIA on helping SK Hynix develop future memory chips. We just don't know. Because we don't know, there's the risk. [00:06:24] Speaker 6: I think Jensen's response to me is basically all of the above, right? And there's no timeline for the deployment of that $500 billion. But SK Hynix has 60% market share in high bandwidth memory. NVIDIA has a greater market share of compute. So that's kind of how it was explained to me. There's the open AI piece of this. And one thing I think the audience really appreciate is there's all the reporting about what NVIDIA is or isn't committed to. But what argument does NVIDIA make historically about whether it is or isn't circular financing? [00:06:54] Speaker 7: Yeah, I mean, it's pretty much what Jensen said to you, which is, look, this thing is going to the moon. It's growing and it's growing and it's growing. I'm making these investments to speed up that growth and to facilitate that growth, make sure it doesn't stall. And guess what? If I'm buying parts of these fantastic companies, then that's going to provide a great return on that investment. That's his argument. But obviously underpinning that is that we're going to the moon and not everybody believes that. [00:07:18] Speaker 6: Right. So it's whether or not you believe in the future state of the market that Jensen's outlined. Just to finish, it was very interesting to be there on Friday. You know, Jensen Wang is still a leader in industry, but it's kind of more than that. You had the president of South Korea, all of the major AI CEOs, and then Jensen kind of leading the way in the middle. What else did we learn from that event? [00:07:39] Speaker 7: Yeah, I mean, you have arguably the two leaders of, you know, the AI software development efforts. You have, you know, Dario and you have Sam Holtman. And we're talking about who's doing what, who's doing chips. Chetaewon is like, oh, well, you know, these guys are making chips now. And that was a revelation. And, you know, Dario was there on stage with him, supposedly having told him this. Jensen, who he would be competing with, was there on stage. So there was a lot of dynamics and that just showed how fluid the situation is. [00:08:08] Speaker 6: On a very serious note, NVIDIA's decline has now accelerated to 5%. Again, markets opened higher. NVIDIA is now down. There's no explicit headline, but the chat in the market is about circular financing. Bloomberg is inking. Thank you very much. [00:08:21] Speaker 8: Joe, to what degree, as an owner of NVIDIA, do you have a handle on, on their various exposures and collaborations and, and where their implied bets are? [00:08:32] Speaker 9: And the reality is that none of us really do have a full handle on it. You mentioned being an owner of NVIDIA. That is a personal position that I've put on in the last several weeks. And I just want to kind of walk everyone through what's going on with NVIDIA. Let's keep something in mind as it relates to moving into the summer months. It's very clear that volumes decline and not so much that liquidity conditions decline, but you see positioning come down. And that's been the story in memory and semiconductors since we went parabolic with Micron's earnings at the end of June. You had an opportunity to build momentum in NVIDIA for the first time, really since the fall of last year, off of the June lows. And it looked, Mike, like we were doing that. We were up to 214 on last Wednesday. We saw a reaccumulation of positioning for people that had stepped to the sidelines, frustrated with the sideways consolidation. That's kind of now neutralized, that momentum that was recognized over the last several weeks. You're back to it being what I would call a yellow light. You're still working against what would be a supportive 200-day moving average down at 192.5 or so. And below that, you have the June lows at 189.80. So this position that I put on, 198, and then an additional purchase up at 208, I will neutralize those positions with a break below 192. And it's very clear that the momentum that we were building over the last several weeks has been neutralized. [00:10:02] Speaker 10: You know, here's what I'd say, that I'd be more concerned about NVIDIA, which is trying to be an indirect recipient of revenue and more business, by funding those like OpenAI or data centers, than I would about Meta or Alphabet or Amazon, who are looking for direct benefit. So we don't know. It's a step removed from what the ROI could be. Yeah. And what they must see from this, the only reason they would do it, is if they see competition, more competition on the horizon, and they're looking to stake out their ground by tying companies up through finance. [00:10:43] Speaker 8: Or if they see the possibility that we get a stutter step in demand for NVIDIA's products because the rest of the world doesn't balks at financing these. [00:10:52] Speaker 10: No, I disagree with that because if they saw a stutter step, then they wouldn't be, then these companies like OpenAI would not need those data centers. So they're not driving. [00:11:03] Speaker 8: No, they're worried about the rest of the world not wanting to finance every single box that these companies want to build. [00:11:09] Speaker 11: You know, I like to think of, from the point of view of free cash flow at NVIDIA, which I'm going to say this calendar year, we know they're on a fiscal year, ends January 31st. But the calendar year, 2026, they're going to generate about $200 billion in free cash flow. The year prior to this, almost $100 billion in free cash flow. These are big numbers. And I ask rhetorically, even though I'm going to answer the question, what are they supposed to do with that free cash flow? Do we want them to start buying back shares and issuing a dividend? Because I'll tell you, that would be a signal to the market that their best days of growth are behind them. I don't want them to do that. Now, we can question, I think it's a reasonable question, are they choosing the right horse in funding OpenAI? I actually have a question to that because of what I said earlier, these Chinese large language models that seem to be doing almost as good as OpenAI codecs, etc. Not quite, but almost with a lot less cap expenditure. Look, what is NVIDIA supposed to do with all this cash? I don't want them to start issuing a dividend. Let them try to pick their winners. And if it becomes a self-fulfilling prophecy because they funded OpenAI, OK, fine. [00:12:10] Speaker 8: Well, first of all, they're hoping that no cash leaves the building, right? This is a guarantee. They hope they don't have to actually be on the hook for it. So the $200 billion of free cash flow is going to still be there. And was Apple a sell in 2013 when they started buying back stock and issuing a dividend? A good rhetorical question. [00:12:30] Speaker 11: And in hindsight, the answer is no, OK? However, at the moment, people said, what are they doing? Sure. When Microsoft issued a dividend? People said about Microsoft, too. Exactly. Because of a slam dunk since then. 25 years ago, whenever it was. It was not looked at favorably. And I don't think it will be looked at favorably when NVIDIA eventually will issue a dividend. [00:12:47] Speaker 8: Jason, the upside leadership today actually is the stuff that's maybe been pressured. It's Microsoft and it's Alphabet to a degree. These have been sort of untrustworthy little rallies, you know, because the charts are a little sloppy. On the other hand, Apple's been, you know, continuing to hit new highs. And it's also leading the way here. [00:13:04] Speaker 12: Yeah, absolutely. One other point on NVIDIA. I look at this deal as almost like it's circular light, right? It's kind of like making sure that they own the infrastructure, backstopping a company like OpenAI to, again, from a credit perspective, ensure that the yield gets done is, does create some leverage for them. Ultimately, it is relying upon execution, which has been the story, like as we follow what's going on with Oracle and the CDS there. So, I think, you know, the market's probably overplaying this one. And this could create an opportunity. And also, to Joe's point of some of the sideways action we've seen over the last nine to ten months, you know, I think this will be an opportunity for NVIDIA in the second half of the year. [00:13:47] Speaker 8: We also have, I mean, Jim's been talking about the Chinese open source models. We also had this IPO, the Chinese memory chip maker. It goes to the moon on day one. And U.S. memory makers are down. I just wonder if this is kind of, look, is more players, more capacity? Yeah. [00:14:04] Speaker 9: Look, I mean, the viewers might think I'm being too simplistic here. But I just think we got reached the point of exhaustion after Micron's earnings in terms of positioning. Let's not forget you had the SK Hynix IPO as well. It's like, you know, how much can actually be allocated in the direction of memory at the expense of the rest of the market? Ultimately, you're making such a significant, concentrated bet towards memory as a portfolio manager. There's a limit to all of that. I think we've reached a limit. I think the market is working off that oversold condition. And the byproduct of that has to be the tape that you have in front of us. There's no other way around the math. [00:14:43] Speaker 13: How about these, you know, in China, CXMT, we've been talking about this IPO for several weeks now. They don't do the highest level, you know, memory stuff, the high bandwidth memory for specifically for AI, but they do a lot of the rest of it. And it's a very tight market. What's your response to the IPO? Is it something you'd consider investing in? Does it disrupt the likes of Micron and other memory names here? [00:15:05] Speaker 14: Yeah, we don't think so. It's not something we're looking at closely right now. Of course, we'll take a look at it as time passes. But I don't think that lower end of DRAM is particularly attractive at this point. You want to stay a little bit more focused on the higher portions. Now, they're going to, China's been making this stuff for some time now. It's just a matter now of how much bigger will they get. And the market itself in China is really quite large as well. So I don't know that it actually has to cross over into the United States. I suspect that they'll have plenty of business in China, probably be very, very focused there for a while. Right now, we're not looking at it. [00:15:37] Speaker 13: That DRAM chart we're showing right now, kind of from April up through now, you can see it going up the hill. It's a round hill memory ETF coming over the hill. It's really moved broadly lower in the past couple of months. How do we finish that chart, you think, into the end of the year? [00:15:54] Speaker 14: Yeah, I think you have to be a little bit patient, but I think we're still higher over the next six months. Remember, if you look back, even at the Sox's retracement, it's been a 25% pullback, largely due to memory. And yet, it's still up 100% in the last year. And earnings, broadly speaking, across semis, were up 120% in that time frame. So we overshot it slightly, and now we've seen the pullback. I think that's normal digestion. I think that probably clears itself pretty quickly. But in the meantime, it's got a lot of folks nervous. It gets a bit of a roller coaster. But ultimately, the demand is still far outstripping the supply for pretty much every memory that the AI is eating into. [00:16:35] Speaker 15: So if the Fed is not the number one most important thing for the equity markets right now, what is? [00:16:41] Speaker 16: Well, I think that the AI trade remains still the most important story. And people, of course, are having longevity doubts. But if someone goes back to 94 to 2000, there were many times when the internet story and even stocks like Cisco came under question whether there was durability. And I think we're in that questioning its durability at this moment. But I think it's still in very good shape. And I think the second big story out there is that margin debt still needs to work off that high level of growth, just like what happened in Korea, which had sort of a margin call. And I think that's why stocks are stalling here. But to me, I think AI still works strongly through year-end. [00:17:23] Speaker 13: Perfect segue. We had this morning, Steve Eisbent was on Squawk Box this morning, the famed investor. Here's what he had to say when he was asked the major question of the earnings season, which is what happens if big tech starts cutting its AI cap-backs? [00:17:36] Speaker 17: NVIDIA, I think when they reported last quarter, had 85% revenue growth. So if the hyperscalers cut, it wouldn't be 85%. And, you know, maybe that would be healthy for the long term. But I think the market would go straight down on that news. [00:17:52] Speaker 13: Just curious for your thoughts, reaction to that. And whether you think that's even likely or worth discussing at this point versus what Google did, which was raise and still get punished. [00:18:02] Speaker 16: Yeah. Well, so on the one level, I'd say I think Steve's logic, there's some logic to it. But the fact that many people are saying that is a sign that we're not at a top because people are questioning the longevity of the cycle. Like, so I think that's actually a bullish thing. The second is, you know, is it probable? I'd be doubtful that there would be the cuts because, you know, these companies still have access to the bond market and the bond market isn't denying them capital. So as you know, CFOs raise money when they can, not when they need to. So I think the spending visibility is going to be very strong. [00:18:42] Speaker 13: Whether it's, you know, even if it's debt funded, you're saying that's just, it's another way that they're going to keep investing. And the longer they keep investing, the longer this cycle goes on, basically. [00:18:51] Speaker 16: Correct. The day CDS markets deny them capital is when then Steve's comments come into play. [00:18:58] Speaker 15: Well, how would we know what would be the trigger? We've already seen credit default swaps and Oracle, for example, go up. Oracle's its own. And by the way, they're not high. They're just higher than they were. I want to make that clear. Oracle's its own thing. What else would you watch as kind of a tip off to that? [00:19:17] Speaker 16: I think if you start to see concessions. So I think that all of these bond raises have been very oversubscribed because we know there's a lot of cash on the sidelines, and these are really good yields, and they're offering good return. When the bond market is not willing to fund, and then you see concessions, that's going to be a sign. And you'll probably see it in the brokers trading weekly, very weak. [00:19:40] Speaker 15: So I think it's not there yet. So almost watch the brokers' stocks as an early potential tell on the AI story. Because I imagine if you wait for the CEO or the CFO of the AI company to say it on the earnings call, by then, the smart money will have already made the trade. Correct. [00:19:59] Speaker 16: Yeah, so the brokers are, as you know, transaction sensitive. [00:20:02] Speaker 18: The supplier companies we are talking about now are much, much stronger than back then. They've got fabulous balance sheets, but the stock sellers, they don't care. They can't stop equating the two periods because the customers are losing gobs here on the spending like they did back then. Now, you may think that NVIDIA guaranteeing $250 billion worth of financing for OpenAI data centers, as it was reported today, makes sense, given that OpenAI is one of their big customers. After all, these chips are insanely expensive, right? If there were no history in these kinds of transactions, you'd think, well, why not? But there is history, boatloads of it, and it is very negative. What we learned in 2000 is that you don't lend to customers who buy your goods. They might default and your earnings get smashed. Smashed. This weekend, we learned of multi-billion-dollar deals where NVIDIA actually makes it possible for the purchase to occur. If the buyer, in this case OpenAI, can actually afford to pay for these chips, perhaps because it becomes public, perhaps because ChatGPT becomes insanely popular, perhaps there are whole new features we don't know about, then NVIDIA's in perfect shape, and NVIDIA stock sellers will look insanely stupid. But if the chip buyer can't pay, well, that's a different story. Isn't OpenAI money good? Why should we worry? Well, first, they haven't come public. Second, we don't know whether they can. Second, they're known to be burning a lot of cash. Third, they aren't considered investment-grade. That makes it much more dicey. The reverberations here are immense, people. There are so many companies counting on the data center for their earnings, all sorts of suppliers. If the market decides it doesn't want to fund any more data centers, not give them more cash, and the companies themselves don't have the money or they don't get paid, then we're back in the year 2000. Why does it resonate so easily? Because in the dot-com era, the companies that bought the goods didn't have enough money to pay for them. Instead, the companies that made the goods ended up on the hook because they provided what's known as vendor financing. And that could be the case right now. Back then, it became a giant game of dominoes, and everyone got annihilated, especially investors in these companies. Could you have seen it coming? Yes, if you paid attention to the balance sheets of the buyers, like OpenAI, when they got grotesque, you had to sell. There's a reason my hedge fund got out of the dot-com stocks about a week before they peaked. The balance sheets of those who were buying supplies, well, they were so bad that they had to rely on the suppliers to afford the goods. That's what could be happening now, and the market hates it, or the stock of NVIDIA would not have been down 10 points. It might have been up 10 points. Yes, you have 10 points today. Notice, I'm not saying NVIDIA doesn't have the money to guarantee sales or to provide vendor financing. They do. I think NVIDIA's balance sheet is among the best in the world. The company's an amazing investor, too. But I lived through 2000, and even the strongest became awful stocks. It took them decades to revisit their previous highs. The stock market is telling you what it thinks of NVIDIA because of these kinds of transactions, even as plenty of people, including me, acknowledging NVIDIA is one heck of a great company. So many of the buyers of NVIDIA AI chips had tremendous balance sheets a year ago. That's no longer the case. Now some desperately need more money to finish their data center build-outs, and it might not be available. Others are losing their investment grade status. That's going to send their stocks lower. Now, let me tell you what sticks in my crawl. Back in 1999, I begged the big suppliers not to do this kind of stuff. I saw it in action, and I pulled out of the stocks. So I cannot sit here and say, don't worry about it, because I was worried then. I have to help you anticipate what sellers, stock sellers, will do. It's not all muscle memory, by the way. My friend, Michael Symblis, my favorite strategist, JP Morgan, he just did an amazing piece last week that compared this moment to the dot-com era. His conclusion? They're too close for comfort. I have to agree. It doesn't matter if your balance sheet's perfect. They all started that way. It doesn't matter if you have bountiful sales. They all started that way. History is brutal. I hope NVIDIA isn't actually making these kinds of transactions. I wish they'd just stay conservative and buy back their own stock. I wish they hadn't given ammo to the short sellers. I wish they'd spend more time thinking about the dot-com collapse. I know they may end up on the hook for nothing. Maybe this time is different. That's what I'm wishing for. But investing isn't about wishing. It's about knowledge. No matter how smart they may be, and they are much smarter than I am. Maybe they never took part in 2000. Maybe they were doing something else. I saw the movie. Heck, I was in a movie. Bottom line. I don't want the sequel. NVIDIA shouldn't make these guarantees, even if it has all the money in the world. Just history. That's all. Just history. We say NVIDIA at the club, don't trade it, because we believe that we will see these kinds of deals aren't worth it. But the big institutions are not going to listen, and they will continue to sell the stock. Why? Because history is on their side. [00:24:46] Speaker 19: All right, I hope you're all doing well today and staying calm in this market. Today was a mixed day throughout the market with a notable divergence among tech stocks. Many tech hardware stocks traded notably lower, while many software stocks traded higher. That divergence is even more evident when looking at a heat map of the NASDAQ 100. There are multiple reasons for Monday's sell-off in semiconductor stocks. There are three main stories. One that's specific to NVIDIA, one that's specific to memory makers, and a third story that's relevant to both NVIDIA and the memory makers. Let's start with the NVIDIA news. Overnight, the Wall Street Journal published a report claiming that NVIDIA is in talks with OpenAI to provide the company with a roughly $250 billion backstop as part of a data center project in Ohio that's being developed by a subsidiary of SoftBank. According to the Wall Street Journal, the project is 10 gigawatts and could cost more than $500 billion. I'm going to interject for a brief moment and say that, if fully built, the 10 gigawatt project would almost certainly cost more than $500 billion. While a 1 gigawatt data center costs $50 to $60 billion today, that number is going to rise closer to the range of $80 to $100 billion per gigawatt, based on Jensen Huang's comments at GTC Taipei in early June. Anyway, back to the story. Earlier reporting in June suggested that OpenAI would control the compute inside the facility under a 20-year lease. And the $250 billion guarantee from NVIDIA covers the data center lease and debt financing, but would not cover the NVIDIA hardware inside the data center. The report claims that NVIDIA is also discussing financing OpenAI's chip purchases up to $350 billion. The first phase of the project is expected to be finished in 2028 with around 800 megawatts of power. And that brings me to a very important point on this story. Even if the report is correct, we're talking about potential financing over the course of many years. If the first 800 megawatt phase is operational in 2028, that's only 8% of the total project. And it's going to be many years before the entire 10 gigawatts is fully built and operational. We're not talking about NVIDIA handing over $250 billion all at once. It will be deployed across multiple financing vehicles and project phases over the course of many years. And so, I do think the market's reaction to this piece of news is an overreaction. Let's keep things in the proper perspective. I think this piece of news had a greater effect on market sentiment than it did on anything fundamental. Some market participants saw this piece of news and immediately assumed that something's wrong with OpenAI. I view it very differently. The leading labs' revenues are surging and their revenues are directly tied to compute. If they had more compute, they would have greater revenues. It really is that simple. Demand is not the problem. The problem is a lack of supply to meet the demand. I wouldn't be surprised if OpenAI's ARR surpasses $100 billion at some point within two years. And with each new generation architecture from NVIDIA, OpenAI's margins should improve as token costs come down. And so, I do think the concerns about OpenAI's spending commitments are somewhat overdone, much like those same concerns were overdone in late 2025. The second piece of news that contributed to Monday's chip sell-off is that overnight in China, memory maker CXMT had a blockbuster IPO, rising more than 400% on its first trading day. After that move, the company is worth nearly $500 billion in market cap, and they raised at least $8.6 billion from the IPO. Market participants expect CXMT to use a large portion of that capital to expand memory capacity. Many market participants are worried about a potential dumping situation in which CXMT floods the market with cheap memory that would challenge the pricing power of the big three memory makers. It's important to remember that just last week, according to multiple sources, Reuters reported that CXMT has been raising prices for months, and even charging some customers more than Samsung and SK Hynix. CXMT would not be raising prices if there was a surplus of supply. Additionally, CXMT is not able to fully satisfy the demand in China, much less the rest of the world, because the demand far exceeds supply. And so, I do think the market's reaction to this news on Monday was somewhat overdone. And now, let's cover the third piece of news that caused both memory makers and NVIDIA to trade lower on Monday. This is the big one, in my opinion. On Monday morning, the information published a story claiming that an unnamed state-backed company in China has begun manufacturing domestically developed immersion deep ultraviolet lithography systems, systems that were previously almost entirely supplied by ASML. ASML traded notably lower on Monday because of this story. Now, I want to provide a disclaimer in saying that this has been an unreliable source in the past regarding NVIDIA China rumors. And so, I don't know if this story is true or not. I'm just bringing it to your attention so that you're aware of it. I also think this story is the main reason why memory stocks traded notably lower Monday morning. According to the story, the unnamed company plans to produce 5 DUV machines in 2026 and 20 machines in 2027. Initial recipients are expected to include memory maker CXMT, which just IPO'd in China, and SMIC. Now, given the expected production volumes of 5 machines in 2026 and 20 machines in 2027, the headline of this story appears to be somewhat sensationalized when it says that China begins quote, mass production. The production numbers in the article seem to contradict the use of the phrase mass production in the headline. The story also claims that China is in the prototype stage of developing a domestic EUV machine. Again, I don't know if the story is true or not. But even if it's true, considering the relatively small expected production numbers and the need for additional reliability testing, I think the market really overreacted to this story Monday morning by selling off memory stocks and other names like NVIDIA. I also think the selling in ASML stock was very overdone on Monday. While EUV machines are not allowed to be sold in China, ASML does sell a lot of DUV systems And the unnamed company in the information story is expected to produce fewer DUV systems in all of 2027 than what ASML sold in China just in Q2 of this year. And so again, I think the market's reaction was really overdone. Let's keep things in the proper perspective. I would also be careful about assuming that the yield from domestically produced DUV systems in China will automatically be comparable with systems from ASML, as that is unlikely to be the case. Overall, I think the market overreacted to these three news stories on Monday. Also over the weekend supply chain, reporting indicated that NAND price increases have begun moderating. It's important to note that this does not suggest prices collapsing, but simply moderating. Reports characterize the market as moving toward a high-priced plateau rather than beginning a new NAND downcycle. This news also contributed to the sell-off in memory stocks on Monday. Now let's cover some more news. On Monday, NVIDIA and Safe Superintelligence announced a long-term partnership to rapidly accelerate SSI's strategic growth. NVIDIA announced they've invested in SSI, but they did not disclose the amount. According to Reuters, NVIDIA invested $5 billion. We don't fully know what SSI has been working on, but NVIDIA entered into this partnership with SSI after obtaining rare access into the company's closely guarded research. SSI was founded by former OpenAI chief scientist Ilya Suitskever. Considering his history and past work, this could potentially turn out to be something very big in the future. Also over the weekend, Jensen Huang spoke with Bloomberg. I'll just mention a few points that stuck out to me. Jensen said that the semiconductor industry is probably going to have to be 10 times larger than it is today over the next decade or so. Jensen said that we're constrained throughout the supply chain. Jensen said these constraints are why he thinks we're going to continue to build out in a throttled way for a decade. Jensen said, quote, I think we have the ability as an industry to double each year, but we're going to have a hard time growing much faster than that. Did you catch that? Doubling each year. If Jensen is saying that the industry has the ability to double each year, what effect do you think that's going to have on NVIDIA's business? As I keep saying in these videos, the total addressable market is growing in the double digits percentage annually. The pie is growing at a strong clip. This is not the time for NVIDIA investors to worry about market share. Jensen also spoke about SK Hynix and the importance of memory, saying, quote, we're going to be purchasing memories from them for many years to come. In order to build a trillion dollars worth of Vera Rubin systems, you're going to have to buy a lot of system memories to go with it. Jensen also made an interesting point about open models, saying that running an open model is not automatically cheaper. Organizations pay for infrastructure, fine-tuning, maintenance, evaluations, and guardrails. The main advantage with open models is control and customization. He indicated that in the future, both open and closed models will be very important. Looking ahead, we have more hyperscale earnings with Meta and Microsoft earnings scheduled for July 29th and Amazon earnings scheduled for July 30th. Overall, I'm expecting each of the hyperscale companies to provide strong guidance and commentary regarding CapEx this earnings season. As for Meta, I'm expecting them to announce strong CapEx guidance. I know there was a bunch of hoopla on July 1st after Bloomberg reported that Meta was developing plans for a cloud business. Some days after that report, Zuckerberg clarified that they do not have excess compute. It's just that some of the deals are very attractive, and Meta could charge a premium if they rented out a portion of their capacity given the constraints throughout the industry. Meta also recently announced they are expanding their Hyperion data center in Louisiana from 2 gigawatts up to 5 gigawatts. Last earnings season, Meta CFO said that they continue to underestimate their compute needs even as they have been ramping capacity significantly. Plus, Meta Superintelligence Labs just recently launched Muse Image, Muse Video, Muse 1.1, and a new model API. Meta is not dropping out of the AI race anytime soon, and I expect their CapEx guidance to be strong. As for Amazon, I'm also expecting strong commentary and guidance regarding CapEx. Amazon CEO Andy Jassy spoke at length last earnings season about Amazon having very high confidence that they will monetize the capacity they're bringing online. As a reminder, AWS is monetizing new capacity as soon as it comes online. Last earnings season, Jassy said, quote, The faster AWS grows, the more short-term CapEx will spend. And then on July 1st, AWS raised GPU rental prices by 20%, and they made that decision based on supply and demand. In other words, demand is very strong and outpacing available supply. As Jassy said last earnings season, the faster AWS grows, the more they will spend on CapEx. AWS is clearly growing. And so I expect strong CapEx guidance from Amazon. Now let's talk about Microsoft, because I think this is the most interesting of the four this earnings season. I want to remind you of a few things. First, Microsoft will be reporting results for the end of their fiscal year. And so they're likely to provide commentary on the earnings call regarding CapEx over the next 12 months. This is going to be a very important earnings call for the entire AI ecosystem. As a reminder, last earnings call Microsoft guided fiscal Q4 CapEx at $40 billion. They also told us that for calendar 2026, they expect to spend $190 billion. Again, that's for the calendar year. And so calendar 2026 would include the third and fourth quarters of fiscal 2026, as well as the first two quarters of fiscal 2027. And so if Q3 CapEx was $31.9 billion, and let's just assume Q4 is $40 billion as Microsoft guided, that leaves $118.1 billion that Microsoft intends to spend in just the first two quarters of fiscal 2027. That would be an average of roughly $59 billion per quarter, much higher than their CapEx so far. What's the reason for that increase? There are two reasons. First, Microsoft is investing heavily in additional capacity for their cloud business. And secondly, Microsoft stated earlier this year that they want to have their own state-of-the-art models in-house by 2027. And they're going to need a lot of capacity to do it. As I said repeatedly ahead of Microsoft's last earnings report, I thought their CapEx guidance was going to be notably higher than what many market participants were expecting. That turned out to be correct. Now, I'll be completely honest. I don't know what they're going to say on the earnings call regarding CapEx over the next 12 months for fiscal 2027. If I had to guess, given the fact that they need additional capacity to compete on cloud, they need to have enough capacity to train their own state-of-the-art models, and also what we're seeing in rising component costs, especially in memory. I think we're likely to get strong next-quarter CapEx guidance. But I just want you to know that market participants' main focus as it relates to CapEx is what Microsoft will say about CapEx over the next 12 months in fiscal 2027. That is what will likely have an impact on the stocks of companies like NVIDIA, Micron, SK Hynix, the NeoClouds, and many others. There is some important nuance in Microsoft's AI strategy, and so we need to listen in to the earnings call to get a better understanding of what's going on. If I could only listen to one earnings call from the four major hyperscale companies this earnings season, I would choose Microsoft's. What they say about CapEx over the next 12 months will likely determine how tech hardware stocks trade the next day. Overall, I'm expecting all four of the major hyperscale companies to report strong CapEx guidance and important commentary regarding AI monetization this earnings season. I don't know what's going to happen in the short term, but from a long-term perspective, I am very confident that NVIDIA will be worth much more in future years than it is today. When Jensen was on the Lex Friedman podcast not that long ago, he was very seriously raising the possibility of NVIDIA becoming a $3 trillion revenue company in the near future. If that happens in the coming years, then it is very plausible that NVIDIA could one day be worth tens of trillions of dollars in market cap. That might sound crazy, but that's what Jensen is implying when he raises the possibility of NVIDIA becoming a $3 trillion revenue company. I guess the question at that point is what multiple the street will be willing to give NVIDIA. I don't know the answer to that question, but I truly do think that NVIDIA will be worth much more in future years than it is today based purely on the fundamental growth of the business. Based on everything I'm seeing, the world is still compute constrained, and I expect that to continue at least through the first half of calendar 2028. In a compute constrained environment, developers will use whatever viable compute they can get their hands on. Today, there are no GPUs that are sitting dark due to a lack of demand, like there was fiber sitting dark due to a lack of demand at the height of the dot-com bubble. Back then, companies were laying fiber in the hopes that use cases and demand would eventually show up. Today, we are seeing the complete opposite. As I've said many times, when market participants compare this AI revolution to the dot-com bubble, they ignore the fact that the internet is already here this time. This means that mass adoption of the technology and new use case development at scale are immediately possible. We don't have to wait years for it to show up. It's already here. The world is compute constrained, which means there is not enough supply to satisfy demand. New capacity is utilized as soon as it comes online. The hyperscalers are monetizing capacity as soon as it comes online. Each of the hyperscalers spoke about being supply constrained on their most recent earnings calls. Additionally, many of the clouds are building out into contracted demand. They're not blindly building in the hopes that demand will eventually show up. No, they're building out because they have signed contracts and, in some cases, significant prepayments from their paying customers. This AI revolution is fundamentally different from the dot-com bubble. In 2026, we'll be a pivotal year for the AI industry thanks to the rapid adoption of agentic AI and the proliferation of agentic systems in the world's leading enterprises. The leading AI labs' revenues are surging right now. Agentic coding and the implementation of agentic systems in large enterprises are new use cases that are increasing inference demand significantly, that subsequently is increasing compute demand. The rapid adoption of agentic AI is why we're seeing an inflection in inference demand. It's why we're seeing the leading AI labs' revenues surge. I wish both Anthropic and OpenAI were public so the public could see the ramp in their revenues. Anthropic's ARR has surpassed $47 billion, up from $9 billion just at the end of 2025. OpenAI is growing rapidly as well. I think the leading labs' surging revenues may be the initial proof point that grabs market participants' attention and causes them to realize that there will be a clear ROI on AI infrastructure. I think the leading labs' surging revenues will also help assure investors of the longevity of NVIDIA's growth, since these labs' revenues are directly tied to compute. If they had more compute, they would have greater revenues. It really is that simple. Demand is not the problem. The problem is a lack of supply to meet the demand. As I've said previously, I expect the world to be compute-constrained at least through the first half of 2028, possibly longer. And so, regardless of what happens in the short term, it's important for long-term investors to remain focused on the fundamentals, maintain a long-term perspective, and remember that we are only in the early stages of egenic systems being adopted at scale. This will increase compute demand significantly. And after that, the next surge in compute demand will likely be fueled by physical AI. We're no longer talking about digital agents performing digital tasks. With physical AI, we're talking about physical AI agents performing physical tasks in the real world. NVIDIA's CFO has called physical AI, quote, a multi-trillion-dollar opportunity and the next leg of growth for NVIDIA. This industry will fundamentally transform society, and NVIDIA has positioned themselves to benefit massively. NVIDIA sells the hardware for the data centers where the models are trained. They offer Omniverse, where the models are taught and tested. And NVIDIA also sells the hardware that allows on-device real-time inference through NVIDIA AGX, allowing robots to have intelligent interactions with the real world even when they are not connected to a data center. Notice that NVIDIA is taking a holistic platform approach to physical AI, and they're embedding themselves as the underlying foundation supporting all of it. Over 2 million developers are already building on the NVIDIA robotic stack, and this is not getting enough attention. As for production ramps, Blackwell Ultra has ramped quickly and remains in high demand. Ruben is on track to launch in 2026. Then we're expecting NVIDIA Grok 3 LPX in the second half of 2026. Later on, we're expecting the launch of Ruben Ultra in 2027 and Feynman after that in 2028. We have a clear data center product roadmap stretching into 2028, and Jensen believes that AI infrastructure spending will reach $3 to $4 trillion annually by the end of the decade. That means Jensen is expecting growing AI demand and an expanding total addressable market underpinning all of this. I don't think we are anywhere near any type of bubble-bursting type of event. With all of this in mind, I seriously think that NVIDIA still has plenty of runway ahead of it, and I think this company will be worth substantially more in future years than it is today. At least, that's my view of the situation. Quick note before I wrap up. All of the compilations on this channel are edited by Finvid with original structure and commentary. Occasionally, the same edits appear elsewhere on YouTube. If you're looking for the original version, it's always here on this channel. Thanks for watching, Finvid. I appreciate your support. Remember to stay calm in this market. Remember to maintain a long-term perspective. And do not make any hasty or irrational decisions. With all of that being said, I hope you all have a great rest of the day, and I'm curious to hear your thoughts about NVIDIA in the comments below. Please leave a like on this video so more people will see it. And while you're down there, please consider subscribing. It's free, and you can always change your mind. Thanks for watching, and hopefully I'll see you in the next video.