If You're Only Buying ONE AI Stock, This Should Be It

There are countless ways to invest in AI right now, but if you only had to buy one stock, this [music] one could be it. Joining us today is MarketBeat analyst Thomas Hughes with a look at one stock that is very new to the market, but he believes has a ton of future potential ahead. Thomas, I'm excited to pour the deep dive into this stock today. Again, this is a really new name, but it is one that is getting a lot of headlines and attention right now. >> Uh yeah, that's uh Cerebras. It's been around since about 2015. It's one of those tech companies whose technology really was established before its time, but now it seems to be pretty important to AI and and that's where the story lies. >> Yeah, when I say this company is new, again, you said it's been around since 2015, but the IPO just happened earlier this year and it's seen a ton of volatility since then. The company just reported its second earnings ever while being a public company and the stock market is not reacting well to that. So, let's get into some of that short-term price action before we move into why you think this is a good long-term play in the AI story. >> Well, right. So, the earnings and the or the revenue and the earnings were both short of expectations, significantly short. So, that really does not uh affirm the analyst expectations, uh provide them a reason to sell, but as you were kind of mentioning, that really is overshadowing the long-term story, which was also strengthened by the results. >> Yeah, looking at this chart to you, does the chart look any different than any other IPO that we've seen this year? >> I mean, it looks pretty standard for an IPO. Um it opened at this high level and it sold off because IPOs typically are are overhyped or overvalued. So, now the market is normalizing and stabilizing. Um it does seem to have established a a bottom in the $200 range and is preparing to move higher, but it could take another few quarters for it to really gain traction. >> Uh one thing we talk about often with uh SpaceX IPO, which is the IPO that most investors are going to be familiar with. One thing that you and Chris have said on this show plenty of times is that you're waiting for the lockup period to end. Uh does anything like that exist with this stock as well? >> That's actually a very a very good question. IPOs typically come with a lockup period. A standard one, unlike most companies and and like SpaceX, this company has got a a staggered lockup expiration. So, the first a tranche was available at the end of the last quarter. Second tranche has been opened up now. There'll be a third and final tranche opened up at the end of the next quarter. Just meaning there's going to be more available shares on the market. Just more liquidity and a reason to think that near term headwinds will exist for the share price. >> Is that staggered lockup a part of the reason we've seen some different volatility points? Or is that volatility just to be expected with any IPO? >> The volatility is to be expected. I think the staggered lockup is maybe accelerating it a little bit because typically IPOs, the lockup is like a a year. You wait for a year after the IPO, the lockup expires and then you can start selling shares. But in this case and with SpaceX, the expiry happened relatively soon and unloaded a lot of shares onto the market. >> So, before we get into the fundamentals of what this company actually does and why you think it's a good long-term play, I do want to talk about that risk factor a little bit as we're seeing the volatility already. Is there a chance that the the bottom could go far lower on a stock like this? Can you as an investor expect to see those IPO highs anytime soon or does it usually take a while to build back to that? >> Um there's always that risk, but it really just depends on on the company itself and and what it's doing and the the revenue trajectory and the earnings trajectory. With this one, we have a very robustly improving revenue trajectory and a relatively quick transition to profitability. So, within that there is a major catalyst ahead just within the next, you know, two to four quarters that could really get this market reinvigorated. >> I think that's where the real skill lies in investing is trying to weed out which ones have real potential and which ones could just be a major flop. And it is always a risk, but there are some of those indicators that you look for that can really make a company stand out. I know that's what you looked for when you made your special report on seven stocks that could be bigger than Nvidia, Google, and Tesla combined. Uh that is also a really interesting list of seven different stocks that have extreme growth potential, which is something that Thomas is really good at looking for. So, if you want to check out that full list, you can scan the QR code or click the link down in the description to get that report for free right now on marketbeat.com. It's normally a $30 fee. Today, it is free for those of you who are watching this video. So, make sure to check out that link. So, Thomas, let's talk a little bit more about this particular stock and why this company stands out to you. What does Cerebras actually do? And what kind of important role does it play in the AI story? >> Well, they're an important piece of the AI story for inference. And inference is the end game for AI. Uh right now, we're building out the infrastructure for AI model building, but when the models are done and they're actually doing their thing, and inference is the output. So, when we put in a query and we get an output, that's the inference. Uh this company is building semiconductor chips that are uniquely uh well suited for inference. Their differentiator is the size of the chip. Uh it's so large that it can house as much memory as it needs on a single chip. And what that really means is like the super super lightning fast speeds, far faster than any other chips can produce because even with the advanced co-packaging we have and with HBM chips, there's still some chip-to-chip latency, connectivity latency, and these giant chips really they eliminate that. There's no latency because it's just all in the same chip. But within that, uh there are some limitations. Uh the chips are really good at spitting out lightning fast results, but they aren't quite so good at doing the thinking part. And that's how they fit into the ecosystem. AMD specifically is working alongside this company, so they have some some native integrations to where AMD's GPUs, its Helios racks, will act as the brain. They'll do the initial memory download. It'll do the thinking and the contextualizing, and then provide the prompts that then go to Cerebras chips, and then they'll spit out some answers. Because they'll be able to load the entire set of data and all the prompts and all the contextualization into one chip, and then just funnel out some answers. And so, what that does for the hyperscalers is provide speed and cost. And both of those things equate to revenue and growth and demand and margins and profitability. >> All right, so we've got another chip maker entering the race, but it does seem like it has differentiator between AMD and Nvidia. Let's talk about that, too. Of There's a lot of different chip stocks out there. We talk about plenty of them on the show all the time. You already talked about how this one is a little bit different, but I want to talk about that competitive scene a little bit. There's so much demand coming in this sector, but there's already plenty of investors who are saying Nvidia's growth story is gone. Lots of people, including yourself, are bullish on AMD's future growth story, but is there still room to grow in this entire chip sector? And let's talk about that growth potential coming here. >> Well, right. And and this is really a a different chip altogether than Nvidia or or AMD. This is not a GPU. This is not like critical to so AI's ability to exist, but it is an enabler of of inference, which is the next stage. And within that, it's a mission-critical component that enables, you know, faster responses at lower cost, and it's fitting well into the MI250 ecosystem. And to me, this is a big catalyst because MI250s are only just now becoming available. The initial deliveries are like this quarter. The ramp will be next quarter and the following year. So, to me, this says that Cerebras real business is still coming ahead. And when we look at the forecast, you think Nvidia sets the standard for how much sales AMD and Cerebras can make. Well, the forecast is for inference infrastructure to be twice the size. So, they can sell twice as many components as Nvidia has. And when we think about the model training, you know, business outlook, it's still really growing. So, generally speaking, the semiconductor market outlook is still just growing. It's got tailwinds. Uh this is a multi-year supercycle that's going to last for I mean, for a long time, I think. >> So, because it's so different, is it an actual direct competitor of Nvidia or AMD or is it in a whole other class that can really work, you know, hand in hand with those two companies? >> It's totally it it's it's a niche play. It's really a competitor for Nvidia cuz it's got its own ecosystem, but it's within the AMD ecosystem and makes the A- AMD ecosystem, you know, more utilitarian and more of a competitor against Nvidia and also, you know, more specifically better for inference, setting it up to capture that big inference market boom. >> Uh is it too closely tied to AMD that really Cerebras' success story is going to be tied to AMD's success story or is there a chance for other partnerships or other use cases outside of just partnering with AMD's chips? >> All right. So, that is really key to the outlook because uh Cerebras is critical to disaggregated inference and that means separating the two parts, the brain and the output, and in that regard, it is being integrated with AMD's system, but it's not going to become vendor agnostic. It could be used with anybody's system. So, as long as inference stays critical, Cerebras should stay critical. >> All right. And that's the question I think that so many people have in investing in AI is will that stay critical because this area is uh developing and unfolding so rapidly and it's it's developing and unfolding in ways that it's very difficult to predict and I think you see that with analyst forecast. You see that with price action in certain stocks. You even see that in earnings and and revenue and when you start to see these backlogs show up. We're not seeing that quite yet with this company. So, let's talk about those actual logistics of how do you know if this is that area of AI where the market is going to explode and really lean into? >> All right. So, the fear of disruption is is real, you know, it could find itself, you know, out in the cold without without any demand. Drivers of that disruption have not yet been built. We have not built out all of the model training facilities. We have not even begun to build out the inference facilities. So, yes, the boom is coming. We've got to build those facilities first to get the outputs that will cause the disruption. So, disruption is possible, but we're talking years in the future. It's not going to be anytime soon. The next year or two will absolutely be strong for this company and you can totally see that in the backlog growth. This is a multi-billion dollar back backlog driven by hyperscale partners. This is a story with legs. >> I know, Thomas, going back to that list that you made of a a full report of seven different stocks that all have that extreme growth potential. They could be bigger than Nvidia, Google, or Tesla, even combined. And so, there are plenty of different stocks to look at in this story. If you want to check out that report again, a reminder to scan the QR code or click the link in the description so you can check out those other seven names. But again, this is something you were always looking for, Thomas, and I think the key is what should people be looking for? Kind of how we started out talking about this video, there are so many options for investing in AI and where it to place your bets, so to say, on what's going to be that next big thing that's going to be needed. Let's dive a little bit more into that, Thomas, and how you select an area that is worth the risk, but also how to play that wisely as an investor. If you think you have a company early like this one that has good long-term potential, how do you weed out all of those names to to to narrow it down to one that you really believe in? >> There's a good example of that in the market today. There are lots of AI critical companies that are making money or are expected to make money and that's the differentiator. Those that are making money today and those that are expected to make money today. There's companies like Nvidia and AMD which are selling hardware. They are absolutely making money today because they don't carry the expense of buying the hardware. So, the neo clouds and the hyperscalers all carry that risk of having to spend the money on the hardware. So, companies that are having to invest in production capability, having to invest in buying materials, companies that are impairing their cash flow or producing, you know, losses are not the great investment today because those are more of a long-term play. But, then companies that are producing the hardware, making sales in that regard, are a better play because they're making money today. Um in the case of Cerebras, uh they are making sales now. They are spending money on ramping their capacity, which is part of why the earnings results were so weak this time. But, that's being uh parlayed into their backlog. So, they're going to ramp capacity by 10x by the end of the year. They're going to start executing on this backlog that's worth decades of of of revenue based on the current quarter. So, their next year forecasts are for triple-digit growth, like 200% growth. It's probably going to be too low. This company is absolutely on track to be exploding into growth, and as it ramps and brings that capacity online, profitability will come. So, next year should be the first full year of profitability with quadruple-digit growth in the years after that. >> That's a huge growth outlook for this company. But, let's talk about what the value of this stock is. Like, cuz I think that that when you're so early, you're you're pre-profitability, um and there's all this volatility in this IPO stock price, it's really just the market trying to figure out what is the value of this stock. If we are going to see that growth in the backlog over the next year and 2 years, how can an investor know, am I getting a fair price for this stock today? What is the value today, and where could it be 2 years down the road? What's your thoughts on on where this this stock stands right now? >> Based on the earnings projections, uh the stock is incredibly highly valued today, and that's because it's just not making um profits today. But, the forecasts are very robust. The P&E falls down into the low teens uh by the early 2030s, and that's based on current estimates. If we see the company continue to build momentum as it has been, ramps its capacity as it plans to do, grows its revenue at triple-digit paces over the next couple of years, uh these forecasts are way, way too low, and we'll see this stock uh continue to rise. Uh it's merely an execution story at this point. It just needs to convert that backlog. Once that starts happening, I think it'll start growing into this valuation, and soon we'll see it uh maybe even command up a premium uh relative to its peers. >> That's definitely a bullish outlook, best-case scenario story for this stock. I do want to just address the other possibility with this one. And And I I say this because you want to look back at some other companies that we've seen that had a lot of excitement around that IPO time, where the IPO'd really well in that first few weeks and months of trade, the stock was trading very, very high, and then it lost so much of that ground because all of that expectation, all of that excitement about what they were going to bring never really executed, never really came to fruition, and the stock lost all of its value, and it never really gained that back. I'm sure you've seen plenty of those stories time and time again, Thomas. So, that risk does seem to be real for this one. >> Of course, it is. I think that EVs are a good example. Those cars Those stocks all IPO'd with high valuations with an expectations for them to do very well. Regardless of what they did, all the government help, they were not able to spark demand for their product. And that's the difference here. There is clear demand for Cerebras' products. It's growing at a very fast rate. Um it underpins a very robust outlook. Um it's fueling the company's ability to expand capacity. Uh so, again, I just bring this back to execution. The company's able to continue executing as it has been, backlog conversion um will become reality, and the stock price will react positively. >> Yeah, that demand story does seem to be the differentiator here with this company compared to EVs, especially. Let's dive into that overall demand in the AI sector a little bit more. Let's talk about what we're seeing with other companies in their earnings report this quarter. Uh And what you're seeing that really shows that that demand is not going away. If anything, it's getting stronger. >> All right. So, not only is the company working with AMD uh to integrate itself into the ecosystem, but there's actual demand from hyperscalers including OpenAI and AWS uh to deploy product. The hyperscale deals are underpinning the outlook and de-risk it to a great a great degree. >> OpenAI, AWS, two huge partners to look at here. Let's talk about constraints potentially. What could slow down the growth of this company? Is there any you know, shortages of products or things out there that could really hold this one back? >> Uh well, that's actually a catalyst for this market. Uh what the chips do and and how they perform, uh they bypass numerous uh constraints and bottlenecks that are tied to the the memory the memory market. So, not only do they not use high bandwidth memory, which is in in great shortage right now, but they also don't require the advanced packaging that HBM products uh need. So, what this means is uh they're they're cheaper to use, they're not constrained by third-party vendor uh limits. They have a cheaper, shorter manufacturing process, and within all that, they've managed to secure capacity at Taiwan Semiconductors to ensure their business through the end of next year well into the 2028. Based on these metrics, it looks like they'll probably be able to sustain that business, you know, long long beyond that. >> Seems like a really good thing for this company, especially when it comes to expense because we know that that's a very costly part of this AI story right now is memory. Uh and we're seeing that you prices continue to increase in that area. So, to not have to be involved in that space of the market um is a good thing for this company. >> Oh, I'm absolutely totally plays into it. It's um it's utility, but also into its cost efficiency um part of the demand by hyperscalers. >> All right. So, that's some really good coverage of this company and what makes it stand out, where the risks are. I'm hoping you can, you know, speak to investors a little bit about whether this is the right investment for them and what you have to know about yourself as an investor to get into a company like this uh where the stock is right now so early in the stage of after it becomes public, uh what kind of investor do you have to be where this is the kind of investment that's going to be the right one for you? And what's the best way to look at investing in a company like this? >> Well, this is still a startup. It's still risky. It's still post-IPO. So, it it takes a certain kind of investor. Buy-and-hold investors can certainly look at this, but you have to to expect volatility. Uh you're going to show losses for sometimes. You might show profits really quickly, but this is one that does carry risk. So, it's one that you're not going to want to put too much money into too quickly, but build a position over time as it uh executes on strategy and presents those bullish catalysts. >> Well, Thomas, thanks for another deep dive into a single stock story. This is a really interesting AI story. If you liked this in-depth look at just one company, we did this with Thomas just a few weeks ago where he talked about Air T and Jewel, and this company just reported earnings, too. Make sure to watch that full interview here.