3 AI Semiconductor Stocks To Love After Their Post-Earnings Sell-Off!!

The demand for AI compute is exploding due to the rapid shift towards agentic AI, heavy inference workloads and also widespread enterprise adoption and usage is shifting from training models to active inference. And the big tech companies, they're projected to spend roughly $750 billion on data centers, advanced chips, and related infrastructure. And that demand surge has led to a capacity crunch. We're seeing higher GP rental prices, hardware shortages, and even service outages across major platforms. That's creating a lot of opportunities for some pretty interesting companies and some pretty interesting stocks. And that's what Jose and I are going to be talking about today. We're going to be breaking down three stocks that we think are built to win from these industry dynamics. >> Hey Jose, how are you doing today? Hey, Rachel. Pretty good. I mean, super excited about today's episode. I know the fools missed me last week because there really wasn't time for me to talk about semiconductor stocks, right? We had big things. So, I'm going to make it up to all the fools to this week. We're going to take a nice look at I'm bringing two semiconductor stocks into the table. Uh Rachel, as you mentioned in the intro, right, the demand for compute continues to rise. Um and and I don't see this changing anytime soon. More importantly, Rachel, the two stocks I'm bringing are two that just reported earnings, and unfortunately, the market is punishing them. Uh, so I actually think there's uh for for me personally, I think there's a an opportunity to keep a closer eye on these two names. The first name I'm going to bring up to is Brocom, ticker AVgo. Now, Brocom just reported earnings this uh early this week. I think on Wednesday, they reported earnings and numbers were insane. Revenue was $29.6 6 billion for this quarter, up 86% year-over-year with their AI semiconductor revenue now making up 56% of total revenue and that segment alone was up 221%. Now guidance guidance for next year they brought up for uh 34.8 8 billion and that's up 93% year-over-year for total revenue and the AI revenue check portion is expected to be up another 236% yearover-year. So the demand is there. They're working with a lot and a lot of great players. For those that are not familiar, Broadcom helps other big players design their own AI chips. They have big customers like Google with the TPUs. Uh right now the TPU V7 is in high volume shipments. The V8 starts product shipments. It started product shipment in quarter three with high volume expected next quarter. Meta is expected to start product shipments next quarter. Jalapeno which is open AI's own AI chip is also reaching tape out which is kind of the step before this massive ramp production. So they have big customers and Thropic is buying a lot of of of TPUs from Barcom at Google that there's a lot of excitement for this play. Now the stock it did take a hit. I I I mentioned the stock unfortunately did take a hit. Last time we were record uh while we are recording the stock is in the mid300s around 350s 340s and the market is fearful of a few things here. One of the things is Lee Bhutan did mention that a lot of portion a big portion of their demand is coming from the big AI companies your open AI your your anthropic and I think right now Rachel the market is not so happy when you have these types of customers being a major backlog right what happens if open AAI fails what happens if anthropic fails I believe if any of those fail some other big player will take up that compute demand I I I just think it would the compute would shift to a different hand and whoever's doing the best intelligence. So, I'm not worried about that, but I see why the market would be. The second thing and I thought was pretty interesting before I kind of send it over to you Rachel was Broadcom mentioned that um one of their limiting factors is the supply chain and one of that part of the supply chain is lasers. Lasers are the demand for lasers far out supply. Uh the other thing is land power and shell. when you're building these multi- gigawatt data centers, you have to make sure you have the land, the power to turn these GPUs online or these AI compute online because the last thing you want is what I would consider dark compute, compute that's just sitting and not being turned on. Uh so they did mention that one of the big issues for some of the forward growth uh in the upcoming years is can these chips be turned on and the fear is if they can't does that build like an overinventory on a lot of these players and does that slow down or create the AI bubble thesis that some people have been talking about. Uh but regardless I think Broadcom great opportunity right now they have a lot of great uh plays. The risk is always there. competition in this AI market, land p Power Shell, the supply constraints. Uh, but the demand is there and we see it with those tripledigit year-over-year growth rates. One of the things that really stuck out to me from the earnings call and I think to a lot of investors was the fact that Broadcom now has a a line of sight as they put it to $115 billion in AI semiconductor revenue in fiscal 2027 and then $230 billion estimated for fiscal 2028. So effectively doubling and then doubling again. Um, another thing as well, you know, CEO Hawk Tan said that Enthropic is on track to overtake Google in 2027 to become Broadcom's largest customer for custom AI chips. OpenAI will move into second place there. And you know, you mentioned this earlier, but unlike Nvidia's kind of general purpose GPUs, Broadcom designs custom AI accelerators that are really tailored to specific AI labs. They're tailored to specific hyperscalers. So, it's more when you look at Broadcom, they're more of an investment on who's locked in as their preferred long-term design partner than, you know, who's making the best chip. And I think that's an important uh distinction. But yeah, these order targets, these custom chip orders, yes, from a handful of hyperscalers and AI labs. And I think that customer concentration you noted obviously is a real risk, but these are also their key customers are some of the most, you know, wellfortified financially speaking companies on earth. And I think you're also seeing that you know yes you have this small number of customers but these are the leaders within the AI revolution that we are seeing these are the companies that are building out a AI infrastructure as we know it and Broadcom is really a direct you know beneficiary of that you know a note on the valuation um Broadcom is rough uh trading about 30% below its 52- week high um and I think that presents you know an opportunity I I'm not necessarily going to go so far as to say that it's trading at a low valuation, but I do think that it does present an undervalued opportunity respective to uh its peers. I think it has a lot of room to run. And I think you look at a business that's, you know, guiding to double their AI revenue this next fiscal year and then double it again in the following year. They're trading well off their highs. I think it's an interesting riskreward setup. I think that this is truly a well-run business and you know they are looking at um incredibly large and growing addressable markets for their AI semiconductor revenue. So you know yes as we get into the 2030s are they going to be able to keep up this growth rate? I think that is still to be determined but they don't have to keep up this growth rate for 5 10 years to be an exceptional business. And so I think there's there's a lot to like about this stock if you're looking uh to benefit from these these tailwinds that we're seeing. >> Yeah, definitely. And um just I mean just to add on to that really quickly before we jump into the second stock, the 115 billion for next year on AI revenue. They mentioned that's just that that's them being conservative, right? The the actual demand is much higher, but they have to take in account to a lot of these supply issues that we talked about earlier on. Um and and and the other thing I >> Yeah, they might be underestimating. >> Um the the other thing is um as you mentioned, right, anthropic and open AI, that's I would say it depends on who you talk to. It could be a bull case or a bare case. I can see why the market would somewhat punish if I I didn't mind Google being your number one customer because I know Google can pay for it. But anthropic being number one, maybe that's where I start to get a little a little iffy on on on the potential revenue growth. Um, but yeah, so so Rachel, now I'm going to pass it over to you so you can give us stock number two before I come back for stock number three. >> Sounds good. So I wanted to talk about Lamb Research, uh, ticker LRCX. This is maybe not a household name. It's probably a company you're familiar with, though, if you if you know the space that Jose and I are talking about today. If you don't Lamb Research, they essentially make wafer fabrication equipment. So that's the the etch tools that chip makers need to build memory chips as well as advanced logic and packaging. And you know when we talk about in other videos companies like Samsung or SKHEX or Micron and that you know race uh to win in the memory industry, all of these companies need Lamb's tools to expand. So, Lamb Research, their revenue is sort of split into two buckets. They have systems revenue, which is kind of the the big ticket equipment that I was talking about, and then they have their customer support business group, and that covers uh spare parts, upgrades, and services tied to thousands of machines that are already installed in fabs around the world. that customer support business group uh recently posted its first ever $2 billion quarter and that's recurring installedbased revenue. So once a fab builds its process around Lamb's etch um and deposition tools, it keeps buying parts and upgrades from Lamb for years. And Lamb sells to essentially every major memory and foundry manufacturer on Earth. You know, I named a few of these. There's also the likes of TSFC, Intel. Um, and you know, shipmakers don't casually, you know, swap equipment vendors. And so the switching costs are a real moat for this business. You know, we talk about competitive modes. This is certainly one. You know, we'll talk a bit about geographic concentration. You know, customer concentration is a concern. Jose and I were talking about for for Broadcom. For Lamb Research, uh, most of their revenue comes from customers in Korea, Taiwan, China, and Japan. The reason for that is that is where uh most of the world's leading edge fabs are concentrated. Obviously, that introduces a lot of growth opportunity, but there's also exposure to trade policy, export control risks. Something to note. Now, they reported their fiscal 2026 uh earnings uh back at the end of July, just in that Q4 of fiscal 2026, they reported revenue of about 7 billion, up 15% year-over-year. their fullear revenue grew 26% to $23 billion. Um this is also a very profitable company. Um they saw growth of almost 40% on the bottom line at the end of that year. And you know where that growth is actually coming from. It's split almost evenly between foundry and memory clients. 44% and 46% respectively. It's a really interesting business. You know, one of the comments that stuck out to me from management on the earnings call, uh, CEO Tim Archer had described an extraordinary setup for wafer fab equipment growth, uh, to borrow his exact phrase, heading into 2027. You know, customers are signaling unprecedented long-term demand. He even raised the 2026 industrywide wafer fab equipment spending outlook to $140 billion. This is also a company that pays a dividend. They recently increased their dividend by about 30%. They regularly return billions of dollars to shareholders. So if you want to think of kind of Lamb in just a research in just a few kind of brief sentences to sum it all up, right? This is a company selling the machinery for the AI boom. And so regardless of which AI chip company or memory maker wins, Lamb Research continues to benefit. So an interesting company to talk about, one we haven't mentioned a lot on recent episodes, so wanted to mention that one in today's. Yeah, equipment players are awesome and and one thing with Lamb Research like you mentioned Rachel is it's not just your TSMC right we're getting this massive demand for AI compute but within that AI compute you're getting this massive demand for memory as well so you have kind of that double-edge that that that double t uh tailwind uh happening if TSMC increases capex that's great because they're going to buy more LAM research and as we continue to have this memory shortage that we might even discuss in another episode um you also have those memory companies like Micron from us SKH Highix like you mentioned um continue to invest in in more and more equipment and as long as this demand for AI compute um and AI chips continue to rise uh uh we're going to see some I think we're going to see some really good numbers and that's the other end of the story right when this cycle ends equipment players tend to feel the pain the most um as a semi bull I don't expect that to happen this year or next year uh but obviously I could be wrong. The other thing that I really I I really want the fools to to kind of keep uh uh keep note of is as Rachel mentioned their service revenue. These equipments are expensive and they make software updates, they make regular maintenance updates and that gets charged on a a yearly basis or a contract basis and that tends to be a really high margins. So, as they're selling this equipment, and I think right now as they have a nice amount of strength on their side, they can go and say, "Oh, if you want this equipment, make sure you sign up for this X-year maintenance program as well." All right. Now, we can jump into stock number three. And this one is a little lesser known company in the semiconductor industry. It's less than $50 billion in market cap. Uh, very similar to Brockcom. It reported earnings and the market punished it and rightfully so. Right. I I do think for for example for Brocom the reasons to get punished make sense. I wouldn't give it that I wouldn't be that bearish on the stock but I can see where the market punished it. Um and the same happens with this stock. So the stock I'm taking a closer look at is Credle, ticker CRVO. They make high-speed connectivity chips and cables uh that help connect GPUs, switches, and memory inside AI data sensors. So they're pretty much kind of that that networking solution. uh they have big partners. Five hyperscalers work with them. Now if we look at the numbers, numbers were insane for revenue. Revenue hit 479 million, up 115% yearover-year. Uh it was up 10% sequentially. So you continue to see that growth. More importantly, guidance guidance for the next quarter was 525 million and it was going to be nearly another 100% year-over-year growth. Again, for the full year, for the fullear, because they're just they just finished quarter 1. For the full year, fiscal 2027, they're estimating at least 85% yearoveryear growth. Uh so massive growth as we're going to discuss. They do have various products in all types of avenues and that's what gets me really excited. But I really want to talk a little bit of why the market is kind of punishing first. Um, one of the main things is customer concentration. Their four their four biggest customers make up about 80% of total revenue, right? So it's right for the market to say how can you what happens if one of those customers leave? Unfortunately, unfortunately, just like we saw with Lamb Research, just like we saw with Brocom, only a selected amount of customers are able to need or and purchase the products that Credo and all these semiconductors have. So, that customer concentration, I believe, comes with the market. Uh the other thing that we ended up seeing is the company used to have about $1.4 billion in cash. Luckily, they have no debt. But that dropped by over 50% this quarter and the market is fearful. Are you investing too much on research and development? Are you making too many acquisitions? One of the things we liked about you was a clean balance sheet. Now, it's it while you have no debt, you cut your cash in half. They made a one-time acquisition here, Rachel, of of a kind of a optics play, and they wanted to pay cash for it instead of raising debt. One-time thing, not a big issue for me. Obviously a yellow flag to keep a closer eye on. Is this balance sheet that used to be so strong um getting uh getting weaker and weaker. Now again that's a yellow flag at the moment not necessarily something I would be too worried about. The other thing is we did see gross margins take a bit of a hit. Um gross margins drop down to about 64% but they did guide that for the rest of the year you expect gross margins around 69 67%. rightfully so, the market is punishing you because research and development is becoming a major part of this AI story and the fear of any dwindle in margins um is is is seen as weakness. It's seen as maybe someone taking market share as making maybe seeing more of like this is becoming now a pricing war opposed to a demand and supply war. Uh so those would be some of the main reasons the stock took a hit. when they reported earnings after hours, the stock was down 20 uh 20%. Now, let me explain why I'm excited about this company. The first thing, like I mentioned, for this year, which is their fiscal year 2027, they're already estimating about 85% year-over-year growth. Now, what gets me more excited is actually next year, fiscal year of 2028 because in fiscal year of 2028, they have various new products coming in. They have silicon photonics designs are coming into play and are ramping in fiscal year 2028. Things like NPL which is near package uh optics is something extremely exciting. Uh they also have active LED cables LLC's which are planned for fiscal 2028. Then you have a lot of memory solutions like their omniconnect which they say revenue begins in fiscal 2028 as well. So you have a huge basket of different segments in the semiconductor industry that are all in some form supply constraint and you're telling me even before that basket comes to the market you're still going to give me 85% year-over-year growth this this year alone. Um so this is a nice high high-risk highreward-based company Rachel I like it. Um but obviously obviously competition is real. You have competitions like Broadcom, like Marll, um like Astera Labs that each touch one part or portion of this business. Uh regardless, the growth opportunity for me outweighs a lot of that risk. >> Yeah, I I think you you hit the nail on the head there. I mean, a couple other things to note. If this is not a company you're as familiar with, this is a business that's gone through quite a transformation in terms of its revenue over the last few years. I mean, they went from about $200 million in annual revenue a few years ago to now north of a billion on a trailing basis and we're seeing management has raised the, you know, guidance for the top line to more than 80% for the coming year. So, this is a company that's run up very fast and a lot of that goes back to the nature of what it does and the customers that it's serving that are operating in this constrained high demand environment. So that sort of one feeds the other. That that's really important to understand is this is not a company that has been growing at this pace up until very very recently. Now you know Jose mentioned the the customer concentration there. Their top top customer is about 30% of their revenue. And I think you know Jose made a great point in terms of the dynamics of customer concentration in the space. It's a risk that we talk about a lot and it is really I think important to bear in mind and build into your thesis if you're investing in any of these companies that we're talking about today or any of their competitors. But that is also just very much the dynamics of this industry. There are a very select number of key players that are integral to the AI buildout. a lot of them, not all of them, you know, profitable businesses, well-fortified companies that are driving this demand that are buying this capacity from the likes of Credo and Broadcom and Lamb and that creates a dynamic where there's customer concentration, but these are also the big players in the space. So, that's something I think you have to be comfortable with if you are putting cash uh into these types of businesses. Uh, but I think that credo is an interesting one. I agree it's a more high-risk, you know, potentially highreward uh play, but I also think it's probably the pick on today's list that is really exposed to some of the physical bottlenecks that we are seeing in the AI industry, you know, right now. I mean, you know, Broadcom profits from more custom chips getting ordered. Credo profits from the fact that once you have all these GPUs in a cluster, getting to them to talk to each other fast enough is a significant engineering problem. and that's something that they help solve. So th those are the thoughts that came to me looking at this business today. One I have on my watch list for sure. Thank you Rachel. And thank you fools for watching this episode. If you have any other picks here in the semiconductor industry, let us know in the comments. Other fools read them as well and it gives us ideas of what else to look at in this space. So take care, have a good day and see you all next