Terafab 21 March 2026

[Daniel Nenni]

I guess if TSMC give capacity to Intel, TSMC will give it to Tesla,
TSMC even accept Crypto current deal with upfront payment deal , LOL
I guess it's terms like payment and long term commitment ,
or maybe this is a way of Elon try to make his bargaining tactics?

TSMC builds fabs based on customer needs. As I said, if Elon writes a big enough check TSMC will build it. I still do not understand why Samsung is not in this narrative. Samsung and Tesla signed a $16.5B agreement through 2033. Is that still on the table? Tesla is also using TSMC, and now they will build their own fab? Exciting times in the semiconductor industry, absolutely.

 

[Daniel Nenni]

Five companies, one fab, and the single point of failure holding it together​

Musk's Chip Gambit

Five companies, one fab, and the single point of failure holding it together

www.airealist.ai

 

[Daniel Nenni]

I have it on good authority that Elon Musk did not approach Intel for any part of this.

 

[Daniel Nenni]

Elon Musk wants to build his own chip factory. The announcement of Terafab—a joint Tesla-SpaceX semiconductor fab in Austin, Texas, targeting 2 nm process technology—is either the most ambitious vertical integration play in decades or the most expensive negotiating tactic in history. Either way, it signals something real about the state of AI chip supply. Here is what Terafab actually means, why the space-grade chip angle matters more than the headline number, and what it tells us about where semiconductor supply chains are heading.

Terafab—Tesla and SpaceX semiconductor factory

Tesla et al. becomes more vertically integrated.

www.jonpeddie.com

 

[Markwrob]

Perhaps Elon will get a reunion tour together for The Fifth Dimension.

"This is the new dawning of the Age of Aquarius..."

 

[blueone]

Perhaps Elon will get a reunion tour together for The Fifth Dimension.

"This is the new dawning of the Age of Aquarius..."

Only three of the original five are alive, so even Musk would have a problem organizing a reunion tour.

 

[MKWVentures]

View attachment 4363

Elon Musk wants to build his own chip factory. The announcement of Terafab—a joint Tesla-SpaceX semiconductor fab in Austin, Texas, targeting 2 nm process technology—is either the most ambitious vertical integration play in decades or the most expensive negotiating tactic in history. Either way, it signals something real about the state of AI chip supply. Here is what Terafab actually means, why the space-grade chip angle matters more than the headline number, and what it tells us about where semiconductor supply chains are heading.

Terafab—Tesla and SpaceX semiconductor factory

Tesla et al. becomes more vertically integrated.

www.jonpeddie.com

Musk has done a lot of things that I thought were impossible.... so he gets the benefit of the doubt. .... but lets start with basics:

If you are building a Fab, you need to know tools, to know the tools, you need a process flow and timeline.

What is the process, what are the tools and what is the timeline for first production wafer out. Musk has to know this or he cannot start the fab.

Side note: Musk could spend what Intel/brookfield did on Fab 52/62 so far (20B total)....... never build a wafer..... implode it..... abandon it .... and it would not have a significant impact on his net worth.

Musk should just buy greenland (give every citizen $1M to declare independence... they will say yes) and lease it to the US. Datacenters will stay nice and cool there.

 

[Daniel Nenni]

I am not sure Musk's play here.
But I am very sure he is very dissatisfied with TSMC's pricing and control. If he wanted 50% of TSMC's capacity, TSMC would definitely not give it to him.

It is beginning to look more like a negotiation move, a head fake. I'm sure there is a method to Elon's madness but it seems to be above my pay grade.

I fully expected Elon to buy a big bunch of Intel stock then announce a working deal with them. That is the Elon I know and love.

 

[blueone]

It is beginning to look more like a negotiation move, a head fake. I'm sure there is a method to Elon's madness but it seems to be above my pay grade.

I fully expected Elon to buy a big bunch of Intel stock them announce a working deal with them. That is the Elon I know and love.

That’s an interesting thought…

 

[Daniel Nenni]

Yesterday, Elon Musk posted on X: “Quantity has a quality all its own. TERAFAB.” Within minutes, the internet fractured into its usual camps—skeptics rolling their eyes, believers leaning in. I found myself in the second group, not because I underestimate the difficulty, but because I’ve spent enough time inside the machinery of this industry to recognize when someone is finally naming the constraint everyone else has quietly accepted.

Terafab doesn’t look like another incremental fab expansion. It is a first-principles reimagining of semiconductor manufacturing from the ground up—and at a scale the world has never attempted: a joint Tesla–SpaceX–xAI facility in Austin, Texas, ultimately targeting a terawatt of compute per year. An initial “advanced technology fab” will iterate chip designs with turnaround measured in days, not quarters. The full Terafab will span thousands of acres and draw more than 10 gigawatts at maturity. Its stated purpose is to supply the exploding demands of autonomous vehicles, Optimus robots, space-based data centers, and frontier AI—chips that must be not only the most advanced in the world, but capable of surviving radiation, running hotter, scaling physically to larger dimensions, and delivering performance no existing supply chain can reliably guarantee.

That last part is the quiet admission that matters most.

Terafab and the End of Semiconductor Complacency

Why Elon Musk’s latest moonshot may be exactly the shock the chip industry needs

sharonhagi.substack.com

 

[Artificer60]

My question is even more basic. Even if you have a process flow and the tools to run the wafers, where do you find the employees to run it? I'm told that Intel had basically started hiring anyone with a pulse as technicians. They even had a class on basic hand tools that taught, among other things, what a screwdriver was and how to use it. I wish I was joking.

 

[Daniel Nenni]

My question is even more basic. Even if you have a process flow and the tools to run the wafers, where do you find the employees to run it? I'm told that Intel had basically started hiring anyone with a pulse as technicians. They even had a class on basic hand tools that taught, among other things, what a screwdriver was and how to use it. I wish I was joking.

Blame the lawyers. I read case law where a technician did in fact jam a screwdriver into his leg and took legal action for lack of proper training.

I do agree with the employment issue here in the Untied States. Robots are great and can certainly run around fabs. In fact, at TSMC they already do but some jobs require people and those people are hard to find.

I'm sure Elon can build a tera fab in 10 years that would be competitive with what TSMC has today (N2). Unfortunately TSMC is not standing still. Neither is Intel Foundry or Samsung Foundry who are also in production at 2nm.

 

[Fred Chen]

View attachment 4364

Yesterday, Elon Musk posted on X: “Quantity has a quality all its own. TERAFAB.” Within minutes, the internet fractured into its usual camps—skeptics rolling their eyes, believers leaning in. I found myself in the second group, not because I underestimate the difficulty, but because I’ve spent enough time inside the machinery of this industry to recognize when someone is finally naming the constraint everyone else has quietly accepted.

Terafab doesn’t look like another incremental fab expansion. It is a first-principles reimagining of semiconductor manufacturing from the ground up—and at a scale the world has never attempted: a joint Tesla–SpaceX–xAI facility in Austin, Texas, ultimately targeting a terawatt of compute per year. An initial “advanced technology fab” will iterate chip designs with turnaround measured in days, not quarters. The full Terafab will span thousands of acres and draw more than 10 gigawatts at maturity. Its stated purpose is to supply the exploding demands of autonomous vehicles, Optimus robots, space-based data centers, and frontier AI—chips that must be not only the most advanced in the world, but capable of surviving radiation, running hotter, scaling physically to larger dimensions, and delivering performance no existing supply chain can reliably guarantee.

That last part is the quiet admission that matters most.

Terafab and the End of Semiconductor Complacency

Why Elon Musk’s latest moonshot may be exactly the shock the chip industry needs

sharonhagi.substack.com

The key understatement from that article: "Terafab will probably not succeed on exactly the schedule first announced."

 

[Barnsley]

It is beginning to look more like a negotiation move, a head fake. I'm sure there is a method to Elon's madness but it seems to be above my pay grade.

I fully expected Elon to buy a big bunch of Intel stock then announce a working deal with them. That is the Elon I know and love.

A lot of big talk going on at the minute