 |
| SpaceX D3 satellite chip image render. |
In the midst of a turbulent year for Tesla and just ahead of SpaceX’s highly anticipated initial public offering, Elon Musk has once again unveiled a project so audacious it seems pulled from the pages of science fiction. The billionaire entrepreneur teased what he calls the Terafab, a chip foundry he boldly claims will be “the most epic chip-building exercise in history by far.”
The announcement, which came via a post on X from SpaceX on March 22, 2026, frames the venture as more than just a new factory. Musk is attempting to tie the fortunes of his three most prominent companies—Tesla, SpaceX, and xAI—into a single, vertically integrated juggernaut designed to produce one terawatt of computing power annually. The joint venture aims to consolidate every stage of semiconductor production, from logic and memory to packaging and testing, under one roof.
But even by Musk’s standards, the gap between the ambition and the current reality is staggering.
A Direct Challenge to TSMC’s Throne
At the heart of the Terafab project is a direct assault on the semiconductor industry’s current order. While Tesla currently relies on established giants like TSMC and Samsung for its chip supply, the Terafab is reportedly targeting 2 nm process technology—the most advanced production node available.
There is plenty of hubris in that ambition. TSMC, the world’s leading contract chipmaker, spent decades and hundreds of billions of dollars building the knowledge base, intellectual property, and intricate supply chain relationships required to reach the 2 nm threshold. It is an ecosystem built on trust and technical precision that no amount of entrepreneurial zeal can instantly replicate.
For Tesla, a company with little background in semiconductor manufacturing, the hurdles are monumental. The specialized machines needed for advanced chip production, such as the extreme ultraviolet (EUV) lithography tools made exclusively by ASML, have multi-year waiting lists. Securing even a handful of these machines would require displacing orders from companies that have been planning their roadmaps for a decade.
The Grand Vision: Orbital Data Centers and Solar Satellites
Musk’s vision for the Terafab extends far beyond powering electric vehicles. He envisions a future where the facility fuels an interplanetary infrastructure. The plan reportedly includes sending 100 million tons of solar energy capture equipment into space annually to power AI satellites. Meanwhile, the computational demands for his projects are astronomical: his Optimus humanoid robots alone are projected to require 100 to 200 gigawatts of chips, while planned satellite arrays demand terawatts of power—a volume that exceeds the output of every current and projected chip manufacturer combined through 2030.
The Murky Math of a $25 Billion Bet
If the engineering challenges weren’t enough, the financial math behind the Terafab is equally murky. Musk is pitching this foundry as a way to secure supply for Tesla and SpaceX ahead of the SpaceX IPO, but the capital expenditure numbers raise red flags.
Tesla’s budget for this year is its largest ever at $20 billion. Yet, the company made less than $4 billion in profit in 2025. The Terafab facility alone carries a price tag of $25 billion—and that is before the inevitable cost overruns that have historically accompanied every Musk moonshot.
To understand the likely trajectory of this project, one only needs to look at the 4680 battery cell program. At Battery Day in 2020, Musk promised 10 gigawatt-hours of production within a year, along with a 50% reduction in costs. Coming back to reality years later, Tesla is at about 2% of that original volume goal, and the 4680 battery is still riddled with price and performance challenges, contributing to issues like the slow charging curve of the Cybertruck.
A Humble Wait for an Epic Scale
Chipmaking is orders of magnitude more complex than battery cell production. The success of this pivot will depend entirely on whether Musk can overcome the significant engineering and financial hurdles inherent in modern semiconductor manufacturing. It is a challenge that has humbled far more experienced players with deeper pockets and longer timelines.
For now, the Terafab and its accompanying orbital data centers represent a compelling vision of the future. But given the laws of physics, the constraints of global supply chains, and the lessons of past production hells, it will likely take many years—if not a decade—to even partially accomplish what Musk has promised. As SpaceX gears up for its IPO, investors will have to decide whether this latest ambition is a stroke of genius or a distraction of epic proportions.