Latest / Black Box Theory Podcast / BBT #16 Snippet | Cerebras IPO Explained
Transcript
- Thomas: So we're going to jump into the Cerebris ⁓ IPO, the AI hardware race goes public. Malcolm: Yeah, man. So just for my folks that may not be familiar with Cerebris, what they do, they recently had an IPO, again, stands for Initial Public Offering. It was one of the largest ever tech IPOs ever. And they are kind of like a direct competitor than Tutu NVIDIA, which is like a pretty tall task considering they're the largest company in the world by a pretty decent margin. Yeah, so they closed their first day trading on Wall Street with a market cap just below 100 billion so that puts it near a few of the companies to close above the mark such as you know meta Alibaba They did close 10 % lower on Friday after his first full day of trading which usually happens with you know IPOs It's usually like a bit of like a rug pool towards the end But when I checked bro, they shot up to me open it like 180 at one point they were trading in like 360 370 a share. Um, so yeah, they uh, Just like some quick information, like what you need to know about Cerebris. I forgot what their ticker is. I believe it's like CBRS. Yeah, CBRS. They make a different type of chip than the classic Nvidia GPU. ⁓ It's like the size of a dinner plate is actually like pretty big. so they ⁓ build the largest chips in the semiconductor industry. And so BigChips, which you need to know, is that they process more information in less time, and they deliver results more quickly. And so up until now. in the video they've been winning the AI chip race because the GPU serve as kind of like the base of, you know, the general purpose workhorses. They're really good at kind of like parallel math necessary for training large language models. But now we're in a new era, bro. We're past chat bots and people just, you know, typing stuff on the chat GPT and getting answers back. We're now more of a agentic kind of powered era where agents and AI are working on behalf of us with a set of directions and they go out and do their own thing. So. Now it's not just training and getting a ton of data. It's about inference, right? So it's about AI being able to understand directions and, you know, natural language processing to go out and do stuff on our behalf without us even a touch of finger. like extensions like Codex and Cloud Cowork and Cloud Code, like all these things that just go out and like just do stuff on our desktop when we're away. This is like, this is where stuff is heading now, right? So while training, so training teaches AI, teaches the AI models to learn from patterns and large amounts of data. inference uses the AI to make data make decisions based on new information. So that's kind like the key difference. ⁓ Yeah, man, it's a, I ain't gonna lie, bro, it's gonna be a crazy it's like this year, this IPO season, this is might be the like I said, not that we've been around for that long, but this is probably the craziest season of IPO and just companies that are coming out. And maybe ever like, like in a very, very long time. ⁓ Thomas: IPO season. Yeah. I think it might be ever man because companies IPOing with these valuations and I gotta say, I gotta say this is it's interesting too. You said they make bigger chips. I thought I know everyone was pushing for the smaller chips, right? So it's interesting to see that they're making chips the size of a dinner plate. And that's what people are looking for because you would think people want compact machinery. You have to think about Malcolm: Bro, Mm-hmm. Thomas: robotics and where we're going in the future, you want that to take up as little space as possible. And that's why thought, know, TSMC was going so big because they keep making smaller chips and, you know, everyone wants a smaller processing with increased power. So the fact that they're making bigger ones, it just makes you think like where, you know, the different uses and applications and I don't know. It's interesting to see that that's there. Malcolm: Nah, no doubt. Just for everyone else, it's just kind of like awareness. So like what me and Tom was talking about here as far as these upcoming IPOs and these crazy companies that are coming out, just to name a few, SpaceX, Elon Musk's company is valued at a whopping $1.7 to $2 trillion in valuation. That's the largest... There are companies that have been around, I'm talking about established companies that have been around for a minute and have not... Micron isn't even a one trillion dollar market cap company yet. There are a lot of companies out there that haven't even touched or cracked one trillion and we're talking about IPO-ing. ⁓ We're talking about a company that's coming out IPO-ing at, you know what saying, at two trillion. So SpaceX's IPO, that would put it above companies or even in the same kind of, like this would be worth more than Broadcom, than Tesla, than Meta, than Saudi Aramco. Samsung, Walmart, Berkshire Hathaway, Eli Lilly, SK Hynex, JP Morgan, Micron, AMD, Exxon, Visa, ASML, Johnson & Johnson, Intel, Oracle, MasterCard, Caterpillar, Chevron, Netflix. You hear the companies I'm naming and these are like very well known, very established companies that have had a history of dominance in their specific sectors. We're talking about companies coming out two-trillion like that. this, like... And that's just one company, right? So we're also talking about we have ⁓ Databricks is coming out. OpenAI, obviously, coming out with their crazy IPO valuation, Anthropic. ⁓ Andrew, Stripe, right? There are like this set of companies, and just like what these set of companies and the technology that they're focusing on in the sector that they're in, this is unprecedented in regard to innovation coming out in one. season all at one time. I've personally never seen anything like this before.