Chip War Tesla Can't Afford to Lose｜Terafab Project EP.1

Kind: captions Language: en If I told you that a uh a cryptic seven-word tweet from Elon Musk is actually >> well a declaration of war to seize control of the global AI chip supply chain. Would you believe it? >> I mean honestly most people would just brush that off as you know typical social media hype. >> Right. Exactly. But we are looking at the structural forces today and it's a very real pivot. >> It is. It's a massive shift. >> So let's unpack this for you. The tweet literally just said, quote, "Terra project launches in seven days." End quote. >> No pictures, no details. >> Nothing. >> Yeah. Nothing at all. Yet, it hit what? Tens of millions of views in 24 hours. >> Easily. The internet exploded. >> Right. So, we've pulled together financial analyses, supply chain reports, and our mission for this deep dive is to figure out why this cryptic message marks Tesla's shift from outsourcing to total vertical integration. >> Well, to really grasp the magnitude here, you have to look at the structural ceiling. Tesla just slammed into. >> Historically, their strategy was a classic outsourcing model, right? They design the chips but hand off the manufacturing. >> Yeah. Like AI4 went to Samsung. >> Exactly. And the first batch of AI5 went straight to TSMC, >> which makes total sense early on. I mean, building your own chip fab is it's incredibly daunting. >> Oh, it's a financial nightmare, >> right? Outsourcing saves a ton of capital. >> But Tesla isn't just selling electric cars anymore. Yeah. When you look at full self-driving, the upcoming cyber cab, the Optimus robots, XAI's Grock, >> they are comput hungry monsters. >> Yes, absolute monsters. >> And Musk even admitted recently that just buying up the available global supply of chips, it's just uh it's not cutting it anymore. >> No, >> the core issue isn't just total global supply though. It's Tesla's place in the pecking order. >> The priority queue. >> Exactly. They are stuck in the production queue right behind tech giants like Apple and Nvidia. >> It's like a waiting line at this super exclusive restaurant, right? >> And Apple and Nvidia are the VIPs who always get seated first. >> That's a great way to put it. >> So Tesla's product roadmap is just permanently trapped by someone else's schedule. >> And waiting for that open slot has severe consequences. If Apple gets the newest, fastest chips 6 months before Tesla does, Tesla's self-driving neural nets are literally learning slower than the competition. >> Wow. Yeah. Efficiency aside, they lose that strategic initiative. >> Totally. So, since they can't get that VIP table, they're just well building their own restaurant. >> Enter Terraab. >> Exactly. We all know Gigafactory solved battery capacity, Terraab jumps a full magnitude higher to solve AI chip capacity. >> Which reminds me of Apple back in 2020. Oh, dumping Intel. Yeah, >> right. They dumped Intel for their own M series chips. They stopped waiting on Intel's release schedule. And Mac performance just skyrocketed >> along with their profit margins. >> Yeah. So, Terrafab is essentially the Apple playbook, but applied to industrial AI. >> If they pull this off, their competitive moat completely transforms. They lock down hardware costs, control the speed of iteration, and you know, guarantee supply stability. >> But wait, I have to push back here. Building a chip fab from scratch is mind-bogglingly expensive. >> It really is. >> I mean, how can they afford tens of billions in capital expenditures without just dragging down their core car business? >> That is the multi-billion dollar question right there. Building a fab is not like building a standard car factory, >> right? >> You need hyper sterile clean rooms, extreme ultraviolet lithography machines, and those cost hundreds of millions of dollars a piece. >> A piece. That's insane. >> Yeah. So, over the next 3 years, we have to closely watch their free cash flow, >> the actual cash they have left over, >> right? After paying for these massive buildings, we need to see if Terrafab is a legitimate new moat or just, you know, a cash burning narrative. >> Which brings us to a muchneeded reality check for you listening. And just a quick reminder, we are strictly geeking out on the business logic in this deep dive. >> Yes, definitely. The above content only represents a deep dive into business logic and does not constitute any form of investment advice. Markets have risks, so please do independent research before making decisions. >> Important disclaimer. And looking at the actual timeline, this March 21st launch, it's almost certainly just a groundbreaking ceremony. >> Right. Just shovels in the dirt. >> Exactly. Real mass production of these chips won't happen until at least 2027. >> Wow. 2027. So, the industry is going to be watching closely to see where that $25 billion actually comes from >> and how TSMC and Nvidia react to potentially losing a massive customer. >> Seriously. So, I want to leave you with this provocative question to chew on. If you were Tesla's CEO facing this massive AI arms race, would you risk tens of billions on the Terrafab expansion? >> Or play it safe, >> right? Or play it safe and focus all your resources on getting the Cyber Cab to mass production.