Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Everyone is analyzing Terafab as a chip factory. I think that's the wrong frame entirely. It isn't a fab. It isn't a memory plant. It isn't Tesla's answer to TSMC. It's a city. Semiconductor manufacturing is the most fragmented supply chain on earth. Lithography, fabrication, memory, packaging, test, each... step sits with a different company, in a different country, with handoffs measured in weeks. Nobody owns the whole stack because nobody can. So if you want to collapse that, you don't build a better fab. You build a place where every specialist can physically move in, bring their core competency, and have it connect directly to everyone else's. Infrastructure designed around the seams rather than around one process step. The tell is who's already involved. Intel is in. Samsung is in. Nvidia is clearly interested. If Micron or SK Hynix show up next, or ASML, the pattern confirms itself. One site, one stack, no handoffs. $TSLA $SPCXshow more

Jo Bhakdi

39,556 subscribers

10,955 views • 19 days ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

11 Comments

Darryn Appleton19 days ago

The ability to understand how supply chains work and how to vertically integrate all the key components of your manufacturing process is what strongly differentiates the Musk companies from all the rest. This is the right read on Terafab. well done Jo. You are indeed a Pioneer.

Gerg Oos19 days ago

wasn’t this clear the night he announced it? wasn’t this practically his description of it? u also left out the part where he said the advanced lab will look at improving processes and new physics. for anyone who’s made an asic, that’s the core dysfunction in the whole industry. it’s effing hard and expensive to try anything different in semicon bc incentives aren’t aligned and it’s not in anybody’s interest to make the industry more efficient. the advanced lab is the key to it all, to close the systemic development loops and get efficiencies no one in the industry has ever bn able to achieve.

🤖🧠👀 AI 4 Everman19 days ago

To set expectations: How many years did it take for Tesla to bring Dry Battery Electrode to scale. (Hint: six) Now do one of the most complicated objects ever created by humans. 👍

Niklas Nelldal19 days ago

Iterations are gonna matter even more in this era

DTDavis19 days ago

"Nobody owns the whole stack because nobody can." Vertical integration is good tax planning *if* the corporation can pull off competency through the whole stack. And Elon has shown he can build vertical stacks that are *better* than competent end to end.

Ross James19 days ago

One of the best analyses I have ever read.

@tnt19 days ago

Wow... nice view!

GSB19 days ago

👍🏻🌟🌟🌟🌟🌟

Shai Machnes19 days ago

First you build TerraFab, then LunaFab then MarsFab. Mars will never be independent if it cannot make its own chips. So you learn on Earth, practice off-earth on the moon, and then the real goal - a sovereign Mars.

Erik Gräfenschnell19 days ago

Intel's latest manufacturing processes are really good for use in space. Intel also was a DRAM manufacturer in a former life ("Only the Paranoid survive") Intel also manufacted FLASH. Intel knows memory (as well)

Future Thesis19 days ago

This is why manufacturing cities like Suzhou, Zhengzhou, Chongqing and Chengdu have done so well. so much of the supply chain, talent and industry is a shirt drive away.

Related Videos

ONE DAY BEFORE THE BIGGEST IPO IN HISTORY, MUSK IS PITCHING A $119 BILLION CHIP FACTORY TO THE ONLY COMPANY THAT CAN BUILD IT. 🇺🇸 This week Musk speaks at ASML's private conference, the firm whose machines print every advanced chip on earth. No ASML, no Nvidia, no TSMC. The topic isn't a secret. It's Terafab, a SpaceX-Tesla-xAI plan for a 2nm fab in Texas. First phase $55 billion, total up to $119 billion. The timing isn't random. Days before SpaceX goes public, Musk is handing IPO investors a growth story beyond rockets: Custom silicon for orbital AI data centers. This is how you justify a $1,750,000,000,000 valuation.

ONE DAY BEFORE THE BIGGEST IPO IN HISTORY, MUSK IS PITCHING A $119 BILLION CHIP FACTORY TO THE ONLY COMPANY THAT CAN BUILD IT. 🇺🇸 This week Musk speaks at ASML's private conference, the firm whose machines print every advanced chip on earth. No ASML, no Nvidia, no TSMC. The topic isn't a secret. It's Terafab, a SpaceX-Tesla-xAI plan for a 2nm fab in Texas. First phase $55 billion, total up to $119 billion. The timing isn't random. Days before SpaceX goes public, Musk is handing IPO investors a growth story beyond rockets: Custom silicon for orbital AI data centers. This is how you justify a $1,750,000,000,000 valuation.

Merlijn The Trader

246,413 views • 4 months ago

Every serious analyst has called Elon Musk's Terafab unrealistic. The CEO of the company that supplies the machines to build it just said it is an upside opportunity and compared it to the largest chip buildouts in history. That is not a small statement. ASML makes the lithography machines that every advanced semiconductor fab on earth depends on. You cannot print a chip without them. There is no workaround. There is no alternative supplier. If the Terafab gets built, every single machine inside it comes from one company. The interviewer asked the ASML CEO directly about one terawatt of compute power. He did not laugh. He did not hedge. He compared it to the massive DRAM megaprojects in Korea and said new projects at that scale are exactly the kind of upside ASML is positioned for. His exact words: "New projects are opportunity." Then he said something that most people missed. He said ASML is already watching the Terafab very carefully. Tracking when things will happen. At what speed. That is not how you talk about a fantasy project. The mainstream take on the Terafab is that Musk is dreaming. The CEO of the most critical company in the entire chip supply chain just said he is preparing for it. Those are not the same position. (Watch the full 11 minutes Interview at Bloomberg TV on YouTube)

Every serious analyst has called Elon Musk's Terafab unrealistic. The CEO of the company that supplies the machines to build it just said it is an upside opportunity and compared it to the largest chip buildouts in history. That is not a small statement. ASML makes the lithography machines that every advanced semiconductor fab on earth depends on. You cannot print a chip without them. There is no workaround. There is no alternative supplier. If the Terafab gets built, every single machine inside it comes from one company. The interviewer asked the ASML CEO directly about one terawatt of compute power. He did not laugh. He did not hedge. He compared it to the massive DRAM megaprojects in Korea and said new projects at that scale are exactly the kind of upside ASML is positioned for. His exact words: "New projects are opportunity." Then he said something that most people missed. He said ASML is already watching the Terafab very carefully. Tracking when things will happen. At what speed. That is not how you talk about a fantasy project. The mainstream take on the Terafab is that Musk is dreaming. The CEO of the most critical company in the entire chip supply chain just said he is preparing for it. Those are not the same position. (Watch the full 11 minutes Interview at Bloomberg TV on YouTube)

Ihtesham Ali

67,385 views • 3 months ago

Culture is genetic because behavior is genetic. This beaver never saw a dam in its life. No beavers or anything else ever taught it to build a dam. It wants to build a dam because it is a beaver. Many beavers together build a big dam. That is beaver culture. Humans are not different. Nothing is different. This is what life is. This is how life works. Your body is your mind. A caterpillar wants to build a chrysalis. A bee wants to build a hive. A lion wants to build a pride. You are not special. You are not above your nature. you are INSIDE of it. The thoughts that we think are genetic thoughts. The crimes we commit are genetic crimes. The art we create is genetic art. Just like this beaver, you can give the animal different sticks and it will build a different dam, but it will always build a dam. And you can give humans different "education," but the human will always use it to do what its genes tell it to do. This is the first big answer that you need. This is the biggest piece of the puzzle. This is how to understand people 90% of the way. You just... notice what they do, and get out of the way, and watch them do it. And if they need sticks, you give them sticks. And if you don't like what they do, you have to get away from them. You cannot train dam-building into them or out of them any more than you can with a beaver. A beaver wants to build a dam because it is a beaver. Whatever you see people build, that's what they wanted to build from the sticks they got in the river they were in. Stop pretending you can change it.

Culture is genetic because behavior is genetic. This beaver never saw a dam in its life. No beavers or anything else ever taught it to build a dam. It wants to build a dam because it is a beaver. Many beavers together build a big dam. That is beaver culture. Humans are not different. Nothing is different. This is what life is. This is how life works. Your body is your mind. A caterpillar wants to build a chrysalis. A bee wants to build a hive. A lion wants to build a pride. You are not special. You are not above your nature. you are INSIDE of it. The thoughts that we think are genetic thoughts. The crimes we commit are genetic crimes. The art we create is genetic art. Just like this beaver, you can give the animal different sticks and it will build a different dam, but it will always build a dam. And you can give humans different "education," but the human will always use it to do what its genes tell it to do. This is the first big answer that you need. This is the biggest piece of the puzzle. This is how to understand people 90% of the way. You just... notice what they do, and get out of the way, and watch them do it. And if they need sticks, you give them sticks. And if you don't like what they do, you have to get away from them. You cannot train dam-building into them or out of them any more than you can with a beaver. A beaver wants to build a dam because it is a beaver. Whatever you see people build, that's what they wanted to build from the sticks they got in the river they were in. Stop pretending you can change it.

hoe_math = PsychoMath

1,192,219 views • 1 year ago

Discovered Materials' Akash Ramdas argues the next chip scaling frontier isn't smaller transistors, it's how many you can stack on top of each other: HBM is already 12 high DRAM going to 16 high "We're starting to do this thing where we're taking a logic chip and then putting memory on top of it. That's where materials can make a difference, because now you're actually thermally limited. The thing that's preventing your scaling is no longer how small is your transistor, but how many transistors can you put on top of each other?" "The high bandwidth memory you'd have seen is a stack of dynamic random access memory. It's a 12-high dynamic random access memory stack right now. One of the roadmaps is to make this even higher, 16-high, and this is how you deal with memory bottlenecks in inference or training." "The next stage of that is putting that HBM on top of a logic die. Both these technologies are going to be partially enabled by new materials."

Discovered Materials' Akash Ramdas argues the next chip scaling frontier isn't smaller transistors, it's how many you can stack on top of each other: HBM is already 12 high DRAM going to 16 high "We're starting to do this thing where we're taking a logic chip and then putting memory on top of it. That's where materials can make a difference, because now you're actually thermally limited. The thing that's preventing your scaling is no longer how small is your transistor, but how many transistors can you put on top of each other?" "The high bandwidth memory you'd have seen is a stack of dynamic random access memory. It's a 12-high dynamic random access memory stack right now. One of the roadmaps is to make this even higher, 16-high, and this is how you deal with memory bottlenecks in inference or training." "The next stage of that is putting that HBM on top of a logic die. Both these technologies are going to be partially enabled by new materials."

MTS

17,566 views • 1 month ago

Founder of lululemon on what he'd tell every 25 year old: "I'd tell them that every person in the world is an individual with a different genetic makeup and a different upbringing and the way that you're thinking is so radically different than every other person in the world and incomparable that if you have an idea and you want to move forward with it, don't worry so much about the competition because nobody will be able to replicate you and the way you think about it."

Founder of lululemon on what he'd tell every 25 year old: "I'd tell them that every person in the world is an individual with a different genetic makeup and a different upbringing and the way that you're thinking is so radically different than every other person in the world and incomparable that if you have an idea and you want to move forward with it, don't worry so much about the competition because nobody will be able to replicate you and the way you think about it."

Sam Stoffel

581,562 views • 4 months ago

Scott on why conviction in an idea is dangerous as an investor but required as a founder: "From the investor seat, I don't think you want to be in love with an idea. That's a risky thing because you'll find a way to express that by investing and make some compromises on the team because you know this idea is just so good, its time has come. But if the team isn't there, 90% of the time it doesn't work. Then there's a whole thing can you swap out the team? But the whole Founders Fund thesis was always no, you need the founder to run the company. But on the company side, you have to be in love with the idea. It's not that rational to start a company. There's a lot easier, more comfortable ways to make money if that's the goal. So it's got to be about what the company is doing specifically. If you want to actually start a company, it better be something that you're really passionate about, or you think that the problem is really important."

Scott on why conviction in an idea is dangerous as an investor but required as a founder: "From the investor seat, I don't think you want to be in love with an idea. That's a risky thing because you'll find a way to express that by investing and make some compromises on the team because you know this idea is just so good, its time has come. But if the team isn't there, 90% of the time it doesn't work. Then there's a whole thing can you swap out the team? But the whole Founders Fund thesis was always no, you need the founder to run the company. But on the company side, you have to be in love with the idea. It's not that rational to start a company. There's a lot easier, more comfortable ways to make money if that's the goal. So it's got to be about what the company is doing specifically. If you want to actually start a company, it better be something that you're really passionate about, or you think that the problem is really important."

Patrick OShaughnessy

33,016 views • 5 months ago

Nobody has war-gamed a global supply chain collapse. Not one government. Not one institution. Why? Because the result is too frightening to put on paper. People imagine the worst case is empty shelves for a few months. That's not the worst case. Globalization isn't a trade agreement between governments. It's embedded in every layer of society. Pull the thread and you don't get a new world order. You get second and third order effects nobody can model. Famine. Political revolution. Fragmentation. There is no bottom to it. The closest historical analogy isn't the Great Depression. It's the Bronze Age collapse. And that took three hundred years to recover from.

Nobody has war-gamed a global supply chain collapse. Not one government. Not one institution. Why? Because the result is too frightening to put on paper. People imagine the worst case is empty shelves for a few months. That's not the worst case. Globalization isn't a trade agreement between governments. It's embedded in every layer of society. Pull the thread and you don't get a new world order. You get second and third order effects nobody can model. Famine. Political revolution. Fragmentation. There is no bottom to it. The closest historical analogy isn't the Great Depression. It's the Bronze Age collapse. And that took three hundred years to recover from.

Predictive History

17,248 views • 3 days ago

Marc Andreessen on the question at the center of data center discourse & why it's so important: "Can you build anything in America anymore? Can you build a factory? Can you build a chip plant? Can you build a power plant? Can you build a refinery? Can you build a pipeline? Can you build housing?" "One of the common themes in American life for the last 30 years is the answer to those questions is generally, no, you can't do any of those things." "Take as an example, Silicon Valley, right? So all the chips are made in Taiwan. Well, 40 years ago, all the chips were made in California." "Can you build things in America? Can you build a factory? Can you build an energy plant? Can you build a data center? Can you build housing?" "And on every single one of those, there's this massive problem which is, right now in many cases, in many places, no you can't." Marc Andreessen 🇺🇸 with Joe Rogan

Marc Andreessen on the question at the center of data center discourse & why it's so important: "Can you build anything in America anymore? Can you build a factory? Can you build a chip plant? Can you build a power plant? Can you build a refinery? Can you build a pipeline? Can you build housing?" "One of the common themes in American life for the last 30 years is the answer to those questions is generally, no, you can't do any of those things." "Take as an example, Silicon Valley, right? So all the chips are made in Taiwan. Well, 40 years ago, all the chips were made in California." "Can you build things in America? Can you build a factory? Can you build an energy plant? Can you build a data center? Can you build housing?" "And on every single one of those, there's this massive problem which is, right now in many cases, in many places, no you can't." Marc Andreessen 🇺🇸 with Joe Rogan

a16z

137,253 views • 4 months ago

Terafab is starting to sound like a lot more than just a chip factory! Tesla and SpaceX are reportedly putting $16.8B into phase one, with plans to handle everything from chip design and fabrication to memory, packaging and testing. The idea is basically to bring pieces of what TSMC, Micron and NVIDIA do under one roof. And the chips could eventually power everything from Tesla cars and Optimus to SpaceX data centers and AI compute in space. $TSLA $SPCX

Terafab is starting to sound like a lot more than just a chip factory! Tesla and SpaceX are reportedly putting $16.8B into phase one, with plans to handle everything from chip design and fabrication to memory, packaging and testing. The idea is basically to bring pieces of what TSMC, Micron and NVIDIA do under one roof. And the chips could eventually power everything from Tesla cars and Optimus to SpaceX data centers and AI compute in space. $TSLA $SPCX

Herbert Ong

31,110 views • 1 month ago

Bridgit Mendler on how Northwood is fixing the space economy's ground infrastructure problem: "The problem that we're solving is for every satellite, it requires a connection point back to Earth so that they can control the spacecraft." "That point of contact on Earth is called the ground infrastructure." "If you don't have it, you don't have a space mission." "It literally is just like a rock in space you're doing nothing with." "There was this whole boom around spacecraft manufacturing." "And interestingly what started becoming the longest pole in the tent was just getting the point of contact on the ground to connect with." "You could build a satellite and launch it faster than you could actually connect with it from the ground." "Which just seemed absurd." Bridgit Mendler with Erik Torenberg

Bridgit Mendler on how Northwood is fixing the space economy's ground infrastructure problem: "The problem that we're solving is for every satellite, it requires a connection point back to Earth so that they can control the spacecraft." "That point of contact on Earth is called the ground infrastructure." "If you don't have it, you don't have a space mission." "It literally is just like a rock in space you're doing nothing with." "There was this whole boom around spacecraft manufacturing." "And interestingly what started becoming the longest pole in the tent was just getting the point of contact on the ground to connect with." "You could build a satellite and launch it faster than you could actually connect with it from the ground." "Which just seemed absurd." Bridgit Mendler with Erik Torenberg

a16z

13,462 views • 6 months ago

🚨WORLD ONLY HAS 2% OF WHAT $SPCX NEEDS Samsung, TSMC, Micron, all added up... NOT ENOUGH. Almost every advanced AI chip is made in Taiwan. One Chinese invasion and say bye bye to them This is a last warning for us And the reason cuts deeper than most realize Musk said it plainly: this is now national security. So he's building the answer. Terafab. 100 million square feet, 50x the Pentagon. He calls it "the most valuable building on Earth." A factory meant to make in ONE year what the whole planet makes in fifty. Connect it to the funny part now :) Remember how Claudflare said AI traffic will hit 1000x more human levels within 5 years? Do you really think Taiwan can back up that demand all alone? 1. Exploding demand 2. Single point of failure in Taiwan 3. ONE company building the next AI chip leader This is the most important building USA will build this decade. Elon Musk is building it BEFORE anyone can even start to think they need it. You really wanna follow me atp...

🚨WORLD ONLY HAS 2% OF WHAT $SPCX NEEDS Samsung, TSMC, Micron, all added up... NOT ENOUGH. Almost every advanced AI chip is made in Taiwan. One Chinese invasion and say bye bye to them This is a last warning for us And the reason cuts deeper than most realize Musk said it plainly: this is now national security. So he's building the answer. Terafab. 100 million square feet, 50x the Pentagon. He calls it "the most valuable building on Earth." A factory meant to make in ONE year what the whole planet makes in fifty. Connect it to the funny part now :) Remember how Claudflare said AI traffic will hit 1000x more human levels within 5 years? Do you really think Taiwan can back up that demand all alone? 1. Exploding demand 2. Single point of failure in Taiwan 3. ONE company building the next AI chip leader This is the most important building USA will build this decade. Elon Musk is building it BEFORE anyone can even start to think they need it. You really wanna follow me atp...

Frogify

21,511 views • 1 month ago

Why Rust is overused right now Martin Odersky (Creator of Scala): "Right now, rust is actually overused because a lot of people push Rust for things higher up in the stack Where a garbage collector is fine, but essentially you still want to code without. And that is, for me, a bit an exercise in doing it intellectually just because you can I don't really think there's a big sense in it. If you have the memory for a garbage collector, you should absolutely use one, because it makes a lot of things simpler. So, yes, of course, if given enough brains, I can write code around and I can do that, but why should you? It can be much simpler."

Why Rust is overused right now Martin Odersky (Creator of Scala): "Right now, rust is actually overused because a lot of people push Rust for things higher up in the stack Where a garbage collector is fine, but essentially you still want to code without. And that is, for me, a bit an exercise in doing it intellectually just because you can I don't really think there's a big sense in it. If you have the memory for a garbage collector, you should absolutely use one, because it makes a lot of things simpler. So, yes, of course, if given enough brains, I can write code around and I can do that, but why should you? It can be much simpler."

Ryan Peterman

215,934 views • 1 month ago

$Elon Musk just posted three numbers that should terrify every semiconductor company on Earth. A terawatt of chips. A terawatt of solar. Ten million tons to orbit per year. That is not a product roadmap. That is a species-level engineering ultimatum. Musk: “Build a terawatt of chips, a terawatt of solar, and 10 million tons to orbit per year.” A terawatt is roughly the entire power generation capacity of the United States. He is not asking for a bigger factory. He is asking for a second grid. The name alone tells you the scale. Terafab. Not gigafab. Tera. A thousand times the prefix. A thousand times the ambition. Every chip company on the planet currently begs TSMC for allocation. They wait in line. They negotiate quarters in advance for a fractional increase in supply. Musk looked at that line and started building the factory that makes it extinct. Vertical integration from lithography to packaging. Design to deployment. Under one roof in Texas. But the factory is not the point. The destination is. Most of this output is not staying on the ground. That is where SpaceX turns from a rocket company into the supply chain for orbital compute. You build the chips. You build the solar. You launch them into the vacuum where the Sun never sets and nothing on Earth can compete. The companies optimizing their server racks in Nevada are solving last decade’s problem with last decade’s ceiling. Musk is fusing Tesla, SpaceX, and xAI into a single organism. One builds the energy. One builds the delivery system. One builds the thing they are both feeding. Nobody has ever welded three companies together into a vertical stack like this. Because nobody has ever tried to do what he is actually attempting. Which is not building a chip factory. It is building the industrial base for a civilization that does not end at the atmosphere. And this is what separates Musk from every other CEO alive. He does not manage companies. He runs them like engineering floors where the only thing that matters is whether you can build. No committees. No twelve layers of approval. No political career tracks disguised as leadership. You either ship or you leave. That is why the most talented engineers on the planet keep walking through his doors. Not because the hours are easy. Because the mission is real and the bureaucracy is gone. Every other company on Earth makes you fight the org chart before you fight the problem. Musk deleted the org chart. Musk: “Join us on this journey.” That is the most understated recruiting pitch for the most ambitious project a human being has ever publicly committed to. And he said it the same way he says everything. Like it is already done.$

Elon Musk just posted three numbers that should terrify every semiconductor company on Earth. A terawatt of chips. A terawatt of solar. Ten million tons to orbit per year. That is not a product roadmap. That is a species-level engineering ultimatum. Musk: “Build a terawatt of chips, a terawatt of solar, and 10 million tons to orbit per year.” A terawatt is roughly the entire power generation capacity of the United States. He is not asking for a bigger factory. He is asking for a second grid. The name alone tells you the scale. Terafab. Not gigafab. Tera. A thousand times the prefix. A thousand times the ambition. Every chip company on the planet currently begs TSMC for allocation. They wait in line. They negotiate quarters in advance for a fractional increase in supply. Musk looked at that line and started building the factory that makes it extinct. Vertical integration from lithography to packaging. Design to deployment. Under one roof in Texas. But the factory is not the point. The destination is. Most of this output is not staying on the ground. That is where SpaceX turns from a rocket company into the supply chain for orbital compute. You build the chips. You build the solar. You launch them into the vacuum where the Sun never sets and nothing on Earth can compete. The companies optimizing their server racks in Nevada are solving last decade’s problem with last decade’s ceiling. Musk is fusing Tesla, SpaceX, and xAI into a single organism. One builds the energy. One builds the delivery system. One builds the thing they are both feeding. Nobody has ever welded three companies together into a vertical stack like this. Because nobody has ever tried to do what he is actually attempting. Which is not building a chip factory. It is building the industrial base for a civilization that does not end at the atmosphere. And this is what separates Musk from every other CEO alive. He does not manage companies. He runs them like engineering floors where the only thing that matters is whether you can build. No committees. No twelve layers of approval. No political career tracks disguised as leadership. You either ship or you leave. That is why the most talented engineers on the planet keep walking through his doors. Not because the hours are easy. Because the mission is real and the bureaucracy is gone. Every other company on Earth makes you fight the org chart before you fight the problem. Musk deleted the org chart. Musk: “Join us on this journey.” That is the most understated recruiting pitch for the most ambitious project a human being has ever publicly committed to. And he said it the same way he says everything. Like it is already done.

Dustin

190,768 views • 6 months ago

Gavin Baker, CIO of Atreides Management made one of the most important and nuanced calls on memory stocks in recent months (Save this). His argument is that based on every memory cycle of the last 25 years, the setup today, prices elevated, sentiment high, supply ramping is textbook time to sell but he adds a critical exception. The one cycle in modern memory history where selling was catastrophically wrong was the mid-1990s, which Baker calls the last true capacity cycle in memory. In that cycle, demand was structurally exploding as the internet era required entirely new computing infrastructure to be built from scratch, and memory had to scale with it in a way that had never happened before. His point is that AI may be that same kind of cycle and not a normal boom bust but a once in a generation capacity buildout where the underlying demand is structural, not cyclical. The reason this argument holds weight is the fundamental shift in what memory is in the AI era. Traditional DRAM was a pure commodity, identical specs, interchangeable suppliers, price determined entirely by supply and demand swings. HBM is the opposite because it is custom engineered to fit a specific customer's chip, co-designed between the memory maker and the GPU designer, with SK Hynix's Vice President literally describing it as shifting from a commodity to a customer-tailored custom business. A single Blackwell Ultra GPU now requires up to 288GB of HBM3E, a 3.6x increase over the H100 and major suppliers like SK Hynix and Micron have already sold out their entire HBM production capacity through the end of the year. Because HBM requires advanced packaging processes like CoWoS that can't be spun up overnight, the bottleneck isn't just wafer capacity but rather runs across the entire manufacturing stack. Bank of America projects the global HBM market grows 58% this year alone to $54.6 billion, and Nomura expects the broader memory sector to nearly double to $445 billion. Long Micron!

Gavin Baker, CIO of Atreides Management made one of the most important and nuanced calls on memory stocks in recent months (Save this). His argument is that based on every memory cycle of the last 25 years, the setup today, prices elevated, sentiment high, supply ramping is textbook time to sell but he adds a critical exception. The one cycle in modern memory history where selling was catastrophically wrong was the mid-1990s, which Baker calls the last true capacity cycle in memory. In that cycle, demand was structurally exploding as the internet era required entirely new computing infrastructure to be built from scratch, and memory had to scale with it in a way that had never happened before. His point is that AI may be that same kind of cycle and not a normal boom bust but a once in a generation capacity buildout where the underlying demand is structural, not cyclical. The reason this argument holds weight is the fundamental shift in what memory is in the AI era. Traditional DRAM was a pure commodity, identical specs, interchangeable suppliers, price determined entirely by supply and demand swings. HBM is the opposite because it is custom engineered to fit a specific customer's chip, co-designed between the memory maker and the GPU designer, with SK Hynix's Vice President literally describing it as shifting from a commodity to a customer-tailored custom business. A single Blackwell Ultra GPU now requires up to 288GB of HBM3E, a 3.6x increase over the H100 and major suppliers like SK Hynix and Micron have already sold out their entire HBM production capacity through the end of the year. Because HBM requires advanced packaging processes like CoWoS that can't be spun up overnight, the bottleneck isn't just wafer capacity but rather runs across the entire manufacturing stack. Bank of America projects the global HBM market grows 58% this year alone to $54.6 billion, and Nomura expects the broader memory sector to nearly double to $445 billion. Long Micron!

Milk Road AI

260,701 views • 3 months ago

someone with your exact Claude plan, is out-producing you three to one it isn't talent and it isn't a secret prompt they built the setup once, so their Claude walks in already briefed yours walks in a stranger every time, the gap is a config, not a ceiling

someone with your exact Claude plan, is out-producing you three to one it isn't talent and it isn't a secret prompt they built the setup once, so their Claude walks in already briefed yours walks in a stranger every time, the gap is a config, not a ceiling

Lagosaur

13,620 views • 2 months ago

Many still don’t understand why Elon is building Terafab Terafab is an extension to all the chip makers in the world It’s not about replacement, not a rivalry and absolutely not competing It’s being built to fulfill the massive chip orders that Tesla, SpaceX and xAI actually need TSMC’s most advanced 2nm capacity is totally booked through 2028 Tesla signed a massive $16.5 billion deal with Samsung back in July 2025 to produce AI6 chips at their Taylor, Texas factory and Samsung is building a Tesla Exclusive chip manufacturing plant to full fill this orders When Elon announced Terafab on March 21, 2026...he made it clear: “That rate is much less than we’d like. We either build the Terafab or we don’t have the chips, and we need the chips, so we build the Terafab” He basically told the chip makers: “Produce as much as you comfortably can. We will take them all. Actually we want even more” Even today Elon said: "SpaceX/Tesla will be always be major customers of TSMC and not competitors in the normal sense of the word" Current production rates are much less than they need....That’s why Terafab exists Terafab is an extension to every chip maker… not competition, not rivalry, absolutely not Even Intel has joined as a partner Even if chip supply improves, massive bottlenecks still exist with memory and advanced packaging You simply can't risk those supply chain breaks at this scale That’s why Terafab is being built to vertically integrate everything - chips, memory, advanced packaging all under one roof, targeting 1 terawatt of AI compute capacity per year This is a ludicrous amount of chips that no chipmaker currently produces at this scale. I don't think even TSMC and Samsung truly understand these numbers yet It’s about building the capacity the future actually demands

Many still don’t understand why Elon is building Terafab Terafab is an extension to all the chip makers in the world It’s not about replacement, not a rivalry and absolutely not competing It’s being built to fulfill the massive chip orders that Tesla, SpaceX and xAI actually need TSMC’s most advanced 2nm capacity is totally booked through 2028 Tesla signed a massive $16.5 billion deal with Samsung back in July 2025 to produce AI6 chips at their Taylor, Texas factory and Samsung is building a Tesla Exclusive chip manufacturing plant to full fill this orders When Elon announced Terafab on March 21, 2026...he made it clear: “That rate is much less than we’d like. We either build the Terafab or we don’t have the chips, and we need the chips, so we build the Terafab” He basically told the chip makers: “Produce as much as you comfortably can. We will take them all. Actually we want even more” Even today Elon said: "SpaceX/Tesla will be always be major customers of TSMC and not competitors in the normal sense of the word" Current production rates are much less than they need....That’s why Terafab exists Terafab is an extension to every chip maker… not competition, not rivalry, absolutely not Even Intel has joined as a partner Even if chip supply improves, massive bottlenecks still exist with memory and advanced packaging You simply can't risk those supply chain breaks at this scale That’s why Terafab is being built to vertically integrate everything - chips, memory, advanced packaging all under one roof, targeting 1 terawatt of AI compute capacity per year This is a ludicrous amount of chips that no chipmaker currently produces at this scale. I don't think even TSMC and Samsung truly understand these numbers yet It’s about building the capacity the future actually demands

X Freeze

53,291 views • 5 months ago

TSMC is exploring a role in Terafab, Elon's Texas chip project. Elon's reply was "Just discussions, but something may come of it." Jeff Lutz 🔋 calls it the battery playbook again. Build at a scale nobody else will, and suppliers can't afford to sit out. Intel's already in, Samsung's at the table, and now TSMC. The likeliest deal has TSMC owning and running the fab, with SpaceX investing or committing to buy chips. C.C. Wei says a new fab takes 2–3 years, plus 1–2 more to ramp. Elon got the leverage before a single wafer shipped. 👀 $TSLA $SPCX

TSMC is exploring a role in Terafab, Elon's Texas chip project. Elon's reply was "Just discussions, but something may come of it." Jeff Lutz 🔋 calls it the battery playbook again. Build at a scale nobody else will, and suppliers can't afford to sit out. Intel's already in, Samsung's at the table, and now TSMC. The likeliest deal has TSMC owning and running the fab, with SpaceX investing or committing to buy chips. C.C. Wei says a new fab takes 2–3 years, plus 1–2 more to ramp. Elon got the leverage before a single wafer shipped. 👀 $TSLA $SPCX

Herbert Ong

216,201 views • 1 day ago

Welfare isn't unpopular, it is a system that we all have to rely on and everyone realises its value. It's unpopular with the billionaire Media who hate every penny you get because it is one they don't.

Welfare isn't unpopular, it is a system that we all have to rely on and everyone realises its value. It's unpopular with the billionaire Media who hate every penny you get because it is one they don't.

BladeoftheSun

38,462 views • 1 year ago

This year, doing something a little different. This is a reminder to everyone that you are your OWN STORYTELLER in your OWN LIFE. No one else can dictate who you are. No one deserves an explanation. I am evolved enough to know that nothing comes easy and no one is perfect. Everyone will judge you. No matter WHAT you do. And/ or what you don't do. Acceptance comes with accepting yourself first. And the difficult journey it takes sometimes to get there. I celebrate everyone taking the time to put one step in front of the next everyday in order to build the next version of themselves for the better. Allowing yourself grace, patience and respect...

This year, doing something a little different. This is a reminder to everyone that you are your OWN STORYTELLER in your OWN LIFE. No one else can dictate who you are. No one deserves an explanation. I am evolved enough to know that nothing comes easy and no one is perfect. Everyone will judge you. No matter WHAT you do. And/ or what you don't do. Acceptance comes with accepting yourself first. And the difficult journey it takes sometimes to get there. I celebrate everyone taking the time to put one step in front of the next everyday in order to build the next version of themselves for the better. Allowing yourself grace, patience and respect...

Christina Aguilera

1,084,348 views • 1 year ago

I BUILT GROK BOT SECOND BRAIN 26 agents, one core, running off a single rule my own memory never followed: capture everything, forget nothing, connect it before you even ask. A voice note on a coffee walk gets split into a task and an idea in under 6 seconds - the core already knows the difference between something you have to do and something you just thought of. An idea from three weeks ago gets linked to the one you just had, automatically, before you remember you'd had it the first time. That visual -the rotating profile, the 26 glowing lines firing out through the skull, each one labeled with the agent running it - isn't a rendered mockup. It's the actual runtime. Every line is a live agent doing its job in real time, right now, in your browser. This isn't a note app. Note apps hold what you give them. This is a nervous system for everything your brain was never built to hold onto in the first place. Full build - the 26 agents, how memory actually persists, what it costs to set one up for real - is in the article.

Aleiah

259,431 views • 1 month ago