Elon Musk has announced Terafab, a semiconductor manufacturing complex that, if completed, would represent one of the most ambitious industrial projects in modern history. A joint venture between Tesla, SpaceX, and xAI, Terafab aims to produce one terawatt of AI computing capacity per year from a facility in Austin, Texas. On April 7, Intel signed on as the primary foundry partner, bringing its 18A process node (a 1.8-nanometer class technology) and a $25 billion initial budget.
The scale of the project is difficult to overstate. Terafab plans to manufacture between 100 and 200 billion custom AI and memory chips annually, starting with an initial capacity of 100,000 wafer starts per month and scaling to one million wafer starts monthly at full capacity. The facility will be split into two main operations: one focused on automotive and robotics chips (for Tesla’s Full Self-Driving system, the Cybercab, and Optimus robots) and another dedicated to AI infrastructure, with approximately 80% of output directed to SpaceX’s planned “AI Sat Mini” satellite constellation.
Musk’s stated reasoning is blunt. He claims that all the world’s existing semiconductor factories produce barely 2% of what his companies will need. Whether that figure is accurate or aspirational, it reflects a genuine strategic calculation: as AI workloads grow exponentially, depending on external chip suppliers creates an unacceptable bottleneck. Terafab is not just an infrastructure project. It is a declaration of technological independence.
For Intel, the partnership represents a lifeline. The company’s foundry division generated only $307 million in external revenue annually, a fraction of what TSMC brings in from its tens of billions in fabrication contracts. Intel CEO Lip-Bu Tan called the Terafab deal essential for the transformation of semiconductor manufacturing in the United States. Intel’s stock rose 4% on the announcement, closing at $52.91.
The deal also carries significant geopolitical weight. If realized, Terafab would establish a third major domestic U.S. semiconductor ecosystem alongside TSMC’s Arizona facilities and Samsung’s Texas operations. At a time when chip supply chains remain a critical national security concern, the prospect of a fully American-built fabrication complex producing cutting-edge processors has drawn attention from policymakers on both sides of the aisle.
But the skeptics have a strong case too. Bernstein Research estimates that actually achieving one terawatt of annual capacity would require approximately $5 trillion in investment, which is over 70% of the annual U.S. federal budget. A single 2-nanometer fabrication facility costs between $25 and $35 billion to build, meaning the announced $25 billion budget covers roughly one facility at partial capacity. The gap between the vision and the initial investment is enormous.
For sales teams tired of cold leads, slow customer responses, and manual processes, Dapta is the ultimate tool.
Dapta is the leading platform for creating AI sales agents specifically designed to increase inbound lead conversion. Respond to your leads in less than a minute with voice AI and WhatsApp that converts.
If you want your team to sell more while AI handles the complex stuff, you have to try it.
For Nvidia, Terafab sends an unmistakable signal. Musk’s companies are among Nvidia’s most important customers, and the decision to build custom silicon rather than continue purchasing Nvidia GPUs reflects a broader industry trend. Amazon has its Trainium chips, Google has its TPUs, and Microsoft is developing Maia. The era of Nvidia’s near-monopoly on AI training hardware may be entering its final chapter, not because a competitor beat them on performance, but because their biggest customers decided to build their own.
Musk has a history of announcing projects of impossible scale and delivering them late but delivering them nonetheless. Tesla’s Gigafactories, SpaceX’s Starship, and the Starlink constellation all faced intense skepticism before becoming operational realities. Whether Terafab follows the same trajectory remains to be seen, but dismissing it entirely would mean ignoring that track record.
The first chips from Terafab are not expected for at least three years, and the full vision of one terawatt per year may take a decade or more. But the strategic intent is already reshaping the semiconductor landscape. When your most important client starts building their own chip factory, the relationship has fundamentally changed.
