…st enough. And you can see in the after hours the stock it was down close to 10%. It's just about 9.3% right now. Down $30, but dipped below $300 a share. We saw today it closed at just about 325 a share. So right now it's actually sub300. I think it's an amazing setup for Google. If you're looking to get in an entry point in Google, if you haven't already, keep your eyes on this stock because if it pulls back any further, guys, it's going to be an incredibly attractive entry point for this stock. Google is a bellweather in this sector. I'm not worried about them at all. I think that they are an amazing entry point sub300. So, that's what I want to take a look at the earnings. There some amazing earnings get dropped on the street. But I mentioned in the beginning, it wasn't just the earnings today. I think it was the revelations that came out of South Korea …
I think it's an amazing setup for Google. If you're looking to get in an entry point in Google, if you haven't already, keep your eyes on this stock because if it pulls back any further, guys, it's going to be an incredibly attractive entry point for this stock. Google is a bellweather in this sector. I'm not worried about them at all. I think that they are an amazing entry point sub300.
Contexte extrait par IA
So right now it's actually sub300. I think it's an amazing setup for Google. If you're looking to get in an entry point in Google, if you haven't already, keep your eyes on this stock because if it pulls back any further, guys, it's going to be an incredibly attractive entry point for this stock. Google is a bellweather in this sector. I'm not worried about them at all. I think that they are an amazing entry point sub300.
Transcription Complète
Guys, welcome back. Well, tonight, four major earning reports just gave us memory investors a remarkable look inside the physical AI buildout. Microsoft and Google disclosed a combined almost 97 billion in quarterly capital expenditures. KLA reported another record quarter and outlined an extraordinary expansion in semiconductor equipment spending. And finally, earlier this morning, Vertive reported that their organic orders more than doubled as demand for data center power and cooling continues to accelerate. But the most important re revelation for today wasn't just the earnings report. It was the fact the news that came out of South Korea and China. SK Heinix said customers are requesting more than twice the amount of HBM capacity that it can supply. And at the same time, Reuters broke a story last night saying that China's CXMT is struggling with HBM3 manufacturing yields. What will that mean for the sector? Is the competition as stiff and as strong as we think in China right now? But I think when you put all those numbers together, it is showing us that the world's leading AI infrastructure buyers are expanding capacity at an extraordinary pace. while advanced memory remains one of the hardest components to manufacture at scale. So tonight, we're going to connect all of the dots. We're going to take a look at the earnings. We're going to look at the equipment spending. We're going to talk about the supply constraints and Micron's latest stock move to determine whether this memory cycle has significantly more room to run. If you haven't done it yet though, make sure you hit like and make sure you subscribe to Market Signal. All right, guys. I want to start off first by just diving into the four key earnings. Now, there were some others today, but I'm highlighting these four that I'm going to pull up in the graphic that I created on the screen for you. And I think you can see I want to talk about Microsoft, Google, KLA, and Vertive. AI buildout, four earning shock waves, Microsoft, and Google as it says on the top, 96.6 billion in quarterly capex. But let's dig through each one. Let's start on the left hand side with Microsoft first. You can see that Microsoft generated 92.3 billion in revenue. It's up about 19% year-over-year. But more importantly, its agents and infrastructure segment. That segment alone reached almost 77 billion, growing at a 25% rate year-over-year. And Azure revenue, that growth was 48%. That was one of the key markers that analysts wanted to see from Microsoft this morning. and they absolutely smashed their earnings when they dropped them on the street. But there are a couple other key numbers in here for us memory investors. The biggest one is capex, the 45.2 billion in a single quarter. It's up 30% but Microsoft is actually being kind of rewarded and congratulated for their discipline in spending because that capex was supposed to be closer to 51 million in the quarter and that had mixed results on the memory sector after that announcement positive for mic for Microsoft but their CFO said they expect next quarter for the capex to be back up over 50 50 billion so I think it's a mixed sign it's good discipline for Microsoft right now but still I'm still very confident about this capex for Microsoft and what it means to the memory sector long time as well and we know that roughly 75% of that capex is directed towards short-lived assets primarily CPUs GPUs and accelerated clusters. Now, if we jump over to the next hyperscaler right next to it, we have to take a look at at Google. And I'm going to show you when I get done this graphic of how the market responded to both Microsoft and Google because guys, it is a divergence of response right now. And there are some reasons for it. So, we'll get to that in a moment. But Google's total revenue hit 127.6 billion. It's up 25%. Their cloud revenue more than doubled to 30.7 billion. It's up over a 100%. 103% year-over-year. That cloud operating income hit 11.12 billion, expanding at about a 36.2% uh they're expanding to a 36.2% operating margin. And to support that demand, Google spent 51.4 billion on capital expenditures. That is a increase of 114% year-over-year. Now what this means for memory investors just those two the hyperscalers earnings we know that every single GPU TPU and uh TPU and server node deployed by Microsoft and Google it requires massive amounts of of advanced memory whether that's HBM3e HBM4 high high density server DRAM enterprise SSDs there are massive amounts of advanced memory that is needed and we're not talking just talking about AI they are committing nearly 100 billion a quarter to the physical hardware that is required to run it as well. And then let's talk about the next two before I show you how the market responded because these are also really important. We're looking at earnings from all four corners of the AI infrastructure buildout with the hyperscalers with the equipment makers and with the you know vertive ability to cool this all this equipment and this and these accelerators down. But KLA they dropped an amazing earnings as well. You can see right in that third column, they reported a Q1 revenue of 4.14 billion. It's up 29% year-over-year. And they guided for the second quarter to 4.7 billion. Management now projects that the global wafer fabrication equipment market will exceed 200 billion next year in calendar 2027. And most importantly for us memory investors or micron investors is that KLA expects memory to represent 40% of its process control systems revenue. And that's just next quarter. They expect DRAM to account for 84% of the memory mix overall. And remember, advanced memory isn't built overnight. You cannot simply flip a switch and produce HBM4. It requires extreme lithography, advanced packaging, and meteorology inspection and meteorology inspection to fix yield issues. And talk about yield issues. I'm going to get to that in a moment with CXMT. So, make sure you stay and watch the video to the end because that's going to be a big piece about the competitive threat. But let's finally finish up this graphic and take a look at Vertive. So, Vertive, we know that they are a huge part of the story as well. all around kind of cool the physical power and cooling infrastructure surrounding these chips. And what they showed us here is that their net sales surged 61% to 4.31 billion. Most importantly though is those organic orders more than doubled. It's 100% growth year-over-year and their adjusted operating profit leaped to 83. It leaped by 83% to 1.09 billion. So the takeaway overall if you look at this graphic is that Microsoft and Google they are buying the compute. KLA is seeing the manufacturing they're seeing the manufacturing equipment needed to build advanced uh DRAM and Vertip is building the liquid cooling needed to keep these centers running. The entire physical pipeline is locked in right now. That's what I took away from these earnings. Let me know in the comments below what you took as took away as well. But I have to tell you, we're seeing a bit of a different response in the market. So I want to show you Microsoft and Google and take a look at what is happening in the after hours after they dropped these prints on the street. You can see the headline here, a market watch for Microsoft. It says the main the major reason Microsoft stock is heading higher after earnings is it said Microsoft showed investors that it's able to grow rapidly while keeping spending in check. It's all around their discipline for capital for capex. It showed right here that Microsoft delivered a combination of growth and profitability that impressed Wall Street on Wednesday, especially against the backdrop of greater scrutiny over artificial intelligence investments. We expected the street expected Microsoft to spend 51 billion on capex. The fact that it came in, just to show you again, it came in at 45 billion. You can see here I'm circling. It's still up 30% year-over-year, but it's less than they expected. Now, their CFO did say they're going to get back up over 50 in the next quarter, but clearly that type of discipline was a positive response in the after hours. We see that Microsoft closed today up about a quarter percentage point, a$116, closed at 532, but in the after hours, it's up 3% right now. Another $16 on top, and getting close to to 5.49 a share. Now, Google, unfortunately, is a very different story. We know their capex was 51, but it wasn't the capex that hurt this. It was, you know, their abil, they said their doubling of cloud revenue wasn't enough to appease Wall Street as investors grow more concerned that AI is hurting the company's ability to make money off of its search queries. Now, this is the case of the whisper numbers or the analyst expectations and Google not meeting those even though they demonstrated significant growth. And just to show you again, like we're talking over here, growth of 25% year-over-year. Their their cloud growth was up 103% year-over-year. So significant growth, but this is being seen as a miss because it didn't hit the analyst expectations. You can see overall revenue for Google grew 25% from a from a year before to 27.6 billion, missing the 128 billion figure expected by analysts. Earnings per share were $212. They were also lower than the analyst expectations of $32 and it was driven by almost a $9 billion paper loss on equity stakes which the company attributed to their investment in SpaceX. So this doesn't concern me with Google. I think you're going to see the stock bounce back pretty quickly. But the the reality is and it's called out right here by this analyst. It says that from Ryan Lee, he said that the results are a bad combination, adding that that Google's continuing their continued cash burn gets harder to justify without monetization or the core business growing to compensate for it. So, it does come back to their investments in artificial intelligence. It's the miss analyst expectations. It's the spending and the fact that they're being perceived as their ability to not monetize it fast enough. And you can see in the after hours the stock it was down close to 10%. It's just about 9.3% right now. Down $30, but dipped below $300 a share. We saw today it closed at just about 325 a share. So right now it's actually sub300. I think it's an amazing setup for Google. If you're looking to get in an entry point in Google, if you haven't already, keep your eyes on this stock because if it pulls back any further, guys, it's going to be an incredibly attractive entry point for this stock. Google is a bellweather in this sector. I'm not worried about them at all. I think that they are an amazing entry point sub300. So, that's what I want to take a look at the earnings. There some amazing earnings get dropped on the street. But I mentioned in the beginning, it wasn't just the earnings today. I think it was the revelations that came out of South Korea and China that also give me strong conviction for the memory sector overall. And if we bring all of these earning reports together into kind of into into and just look at all the results and I think it helps us answer probably one of the single most important questions facing Micron and the semiconductor and other it semiconductor shareholders in general right now. The question is, can global memory supply expand fast enough to satisfy hyperscaler AI demand without sacrificing long-term pricing power? And up until today, the Wall Street bears have argued that new capacity additions and particularly China's domestic DRAM expansion would eventually overupp. But today there was three massive data points that broke that I think that completely they it just completely shatters that bare thesis overall. And the first one I want to take you to is what we saw in Digi Times. It was a comment from the SKH Heinix leadership team. And you could see on the headline SKH says customers want more than twice the HBM it can supply. So right here it says SB SKH Heinix say customers are seeking more than twice the HBM memory volume that it can currently supply as it moves towards shorter memory stacks reflecting supply shortages rather than weakening demand overall. So with them with a massive company like SKH coming out in a public forum and saying that their customers are requesting twice the amount that they can supply in the marketplace. That is a strong signal. We have to pay attention to that signal. That shows us that demand is going to continue to outstrip supply moving forward. And we're seeing customers shift their architecture based on that. They're shifting to smaller stacks overall. As I show you in the article here, shorter stacks reflects supply constraints. SKH said the shift towards fewer DRAM layers per HBM stack is driven by customers seeking to maximize AI accelerator production while memory supplies remain constrained. Reducing stack height allows manufacturers to produce more HBM packages or cubes from a given volume of DRAM output particularly potentially supporting more AI accelerators with the same available available memory supply. So that's a massive signal as I mentioned and at the same time we saw a story break last night that I think I've teased it a few times that I think just challenges the whole competitive threat that we see from China right now. And let me show you the story that Reuters broke last night. It says CX China CXMT to start advanced AI uh memory production despite poor yields. And let's go through some of that the details here. It says they're they plan to start production of HVM 3, but despite they their yield right now is as little as 20% of the chips that make it to the usable amount. That's two per people familiar with the matter said. Now there wasn't any public comment from CXMT or from from SKH Heinix on this story as well. They went to both of those parties. But clearly you can see that CXMT they have sent HBM3 samples to several domestic AI chip designers and it expects to begin limited production over the next two months. yields or which which essentially are the share of chips coming off of the production line that meet specifications are only about 20 to 30% compared with yields of around 70% for SKH the market leader CXMT is also they're pressing ahead despite low yields because Chinese AI chip makers have struggled to obtain enough advanced memory to increase output low yields make production expensive because a large share of those chips must be dis discarded or or reworked overall. So think about what that means. If CXMT is only be able to get a 20% yield in the real world, that is a massive challenge for the company. If CXMT is truly discarding or reworking 70 to 80% of their processed silicon, they are putting that's going to put them at a significant competitive disadvantage behind Micron, Samsung, and SKHix. The article said that in testing their HBM 3E, it performed closer to the older generation HBM3 and because advanced HBM requires vertical stacking or vertical through silicon VAS or TSVs and complex micro bump packaging, achieving 3D stacking efficiently is exponentially harder than building standard DRAM. CXMT is forcing production purely because of domestic Chinese strategic reasons and demand from Chinese companies economically. What this article tells me is they cannot threaten the global supply or the suppliers like Micron, SKhness or Samsung. They can't threaten them at equivalent cost, scale or performance. And like that the article said before, if we look at the architecture changes that we're seeing right now, we're seeing that hyperscalers are adjusting their stacks. They're moving from instead of 12 high HBM stacks to eight high HBM stacks. And that's a crucial architectural detail for memory investors because customers are not stepping down stack height because demand is slowing. They're doing it to solve a physical supply bottleneck that they see in the market. And by shifting to eight high stacks or adjusting the die density, hyperscalers can produce a higher number of total HBM packages out of the same exact volume of DRAM wafer output. This allows chip makers to assemble more total AI accelerators today without reducing long-term memory bit capacity. And here is also why this is even more bullish for the equipment makers like KLA. We reviewed their earnings earlier today because what this tells us is that fewer DRAM layers per stack means that you need more base and more back-end packaging processes to deliver the same memory volume. That explains why KLA is seeing its memory process control mix jump up jump from 27% to 40% next quarter and why they're calling for the total addressable market to reach 200 billion. Now the core takeaway for us memory investors, Micron investors or just anyone in this memory sector quite frankly is when you connect Microsoft and Google's 97 billion in capex. The statement that SKH put out there, the fact that they're seeing a 2x supply deficit, KLA's equipment explosion and CXMT's 20% yield wall, it proves that the packaging and manufacturing moat around advanced DRAM is just as wide as ever. The physical AI buildout isn't just expanding. It's running directly into a structural HBM supply wall. And that supply discipline that protects Micron, Samsung, and SKH's elevated gross margins well into 2027 and 2028. And guys, I want to jump over and just take a look at how Micron performed today, what it's doing in the after hours as well before we wrap up the video. We know we had a strong day for Micron today. You see it's up $54, you know, just about 54.5. It closed at 9.83 a share. It's up almost 6%, so five 5.86%. And it had traded on heavy volume today. There was 30.6 million shares. The stock traded as high as 990 if you look to the right in the intraday overall as capital kind of continued to flood into the memory equities following that massive earnings that SKH dropped on the street yesterday and their guidance around shortages being out to 2030. Now, here's the thing. Samsung has probably dropped their earnings 18 19 minutes ago. They were supposed to drop it at 900 p.m. Eastern. I started this video a little early, so we have to keep our eyes on that. actually want to take a look over here in real time. We see Samsung's earnings down right now in comparison to SKINX. If you look to the right, so I'm curious. I'm going to dig into that after this video. And if there's any major updates, I'll drop it onto the channel for you for you guys to know. But if we look at the after hours right now, you see that Micron is down by about two and a half%. It's trading right now at 950. We'll call it 958 a share. But don't let this short-term after hours volatility confuse you or concern you or or worry you at all. Here is why we're seeing an after hours drop in my opinion and what I expect to see at tomorrow's opening bell. I think right now, you know, both with Google and what we saw with Meta last week, there is just this after hours drag. Google shares have dipped in the after hours trading due to that concern around capex, their concern around their ability missing those analyst expectations, their ability not to monetize fast enough overall and it's dragging down the broader tech sentiment momentarily. I think the other piece is we do see traders are probably taking a little bit of quick profits ahead of SanDisk's earnings tomorrow as well as Amazon's earnings tomorrow. But as we get into the opening bell tomorrow with Micron, I expect to see a little bit of initial chop for the stock overall. I think we could see it kind of hover around that 960 970 support zone. We got to look to see if it can get to that 960 970 support zone at the opening bell as market makers absorb the after hours options rebalancing. However, given that Microsoft and Google just confirmed 97 billion in capex and KLA sees a 40% memory mix and the Chinese HBM supply remains completely crippled by a 20% yield, the structural bid underneath Micron remains exceptionally strong. We continue it every time we see these earnings, we continue to see the physical the physical investment cycle supporting HBM server DRAM and enterprise SSDs expanding across every single layer of the supply chain. Hyperscalers are funding it, equipment suppliers are building it, and cooling providers are enabling it. So guys, let me ask you a question. Do you think Micron breaks $1,000 before the end of this week? Drop your comments. Let me know. Let me know what your takeaways were for today or if you think we'll break that $1,000 mark. Let me know your thoughts. If you haven't done it already though, make sure you hit like and make sure you subscribe to the channel. I'll see you on the next.
Commentaires 0
Connectez-vous pour rejoindre la discussion.
Se connecterAucun commentaire pour l'instant. Soyez le premier à partager votre avis !