Elon Musk announced that Tesla's Terafab project, a plan to build what he envisions as the world's largest semiconductor fabrication plant, will launch within seven days. The announcement, made via social media on 14 March, marks Tesla's formal pivot from chip design to chip manufacturing. It also represents a moment of genuine corporate candour: Musk is essentially telling the world that his company no longer trusts existing suppliers to meet its needs.
This is not hyperbole. Tesla first confirmed Terafab on its January 28, 2026 earnings call, where Musk told investors the company needs to build a chip fabrication facility to avoid a supply constraint projected to materialise within three to four years. The calculus is straightforward. As Tesla scales production of autonomous vehicles, humanoid robots, and AI-powered systems, external foundries cannot guarantee the volume or timeline the company demands. The project is projected to produce 100–200 billion AI and memory chips annually, targeting 100,000 wafer starts per month, at an estimated cost of approximately $25 billion.
The ambition is staggering. At full scale, Terafab's ambition of 1 million wafer starts per month would represent roughly 70% of TSMC's current total output, from a single US-based facility. Tesla is targeting 2-nanometre process technology and anticipated to be the most advanced node currently in commercial production. This is not a manufacturing expansion; it is a bet that Tesla can do something no private company outside Taiwan and South Korea has managed: build world-class semiconductor capacity from scratch.
Here's why the scepticism matters. The problem is that Tesla has absolutely zero experience manufacturing semiconductors, and its track record in the closest comparable venture, building its own battery cells, should give anyone pause. Tesla's decision to get into chip manufacturing echoes a similar strategic move from six years ago, when the company unveiled its 4680 battery cell at Battery Day in September 2020. Back then, Musk made bold promises: Tesla would build 100 GWh of in-house cell production capacity by 2022, cut battery costs by 56%, and use the savings to deliver a $25,000 electric vehicle. None of that happened on schedule.
The semiconductor industry has responded with barely concealed alarm. Jensen Huang, who knows more about the semiconductor supply chain than almost anyone alive, is telling him it's "virtually impossible", and Huang has every incentive to want more chip production capacity in the world. A modern 2nm fab requires mastery of extreme ultraviolet lithography, materials science at the atomic level, yield management across hundreds of interdependent process steps, and a trained workforce that simply does not exist at scale outside of Taiwan and South Korea. TSMC has spent forty years and hundreds of billions of dollars building that expertise. It still loses yield on new nodes for the first six to twelve months of production.
What deserves credit, however, is the strategic clarity. For years, Tesla has designed its own processors while outsourcing their manufacture to Asian foundries. That arrangement has become a bottleneck. As global demand for AI chips has surged, Tesla has found itself competing for production slots alongside Apple, Nvidia and AMD, with wait times stretching to unprecedented lengths. Musk eventually concluded that even the most optimistic production scenarios at TSMC or Samsung would not meet Tesla's needs, and so the idea of building the factory himself took shape.
The capital requirements are sobering. Tesla has already outlined plans to invest $20 billion in capital expenditures for 2026, representing more than double the previous year's spending of $8.53 billion, which itself reflected a twenty-five percent decrease compared to 2024. Notably, the $20 billion dollar forecast does not include expenses related to the Terafab project or a potential solar manufacturing facility, suggesting that actual capital requirements could substantially exceed current projections and potentially impact Tesla's cash reserves. During the first quarter earnings call, Tesla CFO Vaibhav Taneja indicated that the company might pursue borrowing to finance its extensive capital expenditure requirements.
The project reflects something deeper: a company attempting to control its entire supply chain in an era of geopolitical instability. By keeping its semiconductor manufacturing process in-house, Tesla can significantly minimise the risk of intellectual property theft, which remains a persistent issue in international partnerships. This decision also aligns with current political sentiments in the U.S., aiming for long-term strategies toward tech sovereignty and resilience. That rationale is persuasive. It is also, by definition, a concession that existing global structures cannot be trusted.
Whether Tesla succeeds is the only question that matters, and a Terafab launch in seven days almost certainly does not mean a fully operational chip fabrication facility opens its doors. Semiconductor fabs of this scale take years to construct and commission. What launch more likely signals is one of the following: a formal project announcement with location and timeline details, a groundbreaking ceremony, a public reveal of the facility's design and specifications, or the start of construction on the first phase.
The underlying tension is unresolved. Tesla has earned genuine credibility on vertical integration; the company built battery production capacity and has now designed competitive chips. But semiconductor manufacturing is different. It is not entrepreneurship or industrial reorganisation. It is physics at the atomic scale, perfected through decades of institutional knowledge. Musk's conviction that money and first-principles thinking can compress any timeline collides with an industry that has proven far more resistant to disruption than automotive or energy storage.
Success would be transformative. Failure would be an expensive lesson in the limits of vertical integration. The truth, as always, lies somewhere in between.