Intel teams up with Musk's Terafab as a foundry partner in a $25 billion chip megaproject.

      In summary: Intel has become the main foundry partner for Elon Musk’s Terafab, a $25 billion joint venture involving Tesla, SpaceX, and xAI that aims to generate a terawatt of AI computing power annually. This partnership provides Intel, which has been facing challenges, with the prominent client it has been seeking since shifting to a foundry-first approach.

      On April 7, 2026, Intel announced its collaboration with the Terafab project, marking its role as the foundry partner for what is regarded as the most ambitious semiconductor facility proposed in the United States. This announcement followed Musk’s initial presentation of Terafab at Giga Texas in Austin, where it was revealed that the initiative plans to deliver one terawatt of AI computing each year. Intel is expected to leverage its advanced process technology, packaging expertise, and manufacturing scale to turn this vision into reality. For Intel CEO Lip-Bu Tan, this agreement represents the most important external partnership achieved since he took office, as he has been focused on rebuilding Intel's foundry business.

      What Terafab aims to develop

      Terafab is envisioned as a comprehensive semiconductor complex that will integrate chip design, lithography, fabrication, memory production, advanced packaging, and testing within a single facility. It aims to produce between 100 billion and 200 billion customized AI and memory chips annually. The initial construction aims for 100,000 monthly wafer starts, with plans to eventually reach one million per month at peak capacity. The project will utilize two distinct facilities on the Giga Texas campus: one for automotive chips and humanoid robotics applications, including Tesla’s Full Self-Driving and Cybercab programs, and the other for high-performance AI data center infrastructure and specialized processors for orbital missions.

      The orbital aspect is critical to the project's vision. SpaceX, which completed its acquisition of xAI in February 2026 for a merger valued at approximately $1.25 trillion, is developing a constellation of space-based AI satellites known as AI Sat Mini. Musk indicated that 80% of Terafab's computing output will support this orbital infrastructure, while the remaining 20% is intended for ground-based uses. Although the total cost of Terafab is estimated between $20 billion and $25 billion, independent analysts have expressed significant skepticism about the adequacy of this amount to meet the ambitious production goals. Bernstein Research has suggested that the actual capital required to achieve one terawatt of annual computing could be as high as $5 trillion, exceeding 70% of the total annual US federal budget.

      Intel’s contribution and the value of the deal

      Intel will provide its 18A process node, the company’s leading logic manufacturing technology, which is currently being prepared for high-volume production at its fabrication facilities in Arizona and Oregon. Intel’s 18A represents a 1.8-nanometer-class node, aligning it with the most advanced global commercial production processes, and showcases the highest level of semiconductor technology manufactured entirely in the United States. Intel's statement on its involvement with Terafab was straightforward: “Intel is proud to join the Terafab project with SpaceX, xAI, and Tesla to help redefine silicon fabrication technology.” Additionally, the company stated, “Our ability to design, fabricate, and package ultra-high-performance chips at scale will accelerate Terafab’s goal of producing 1 terawatt per year of computing to drive future advancements in AI and robotics.”

      Tan’s post on X presented a more personal perspective, stating, “Elon has a proven track record of reimagining entire industries. This is exactly what is needed in semiconductor manufacturing today. Terafab represents a step change in how silicon logic, memory, and packaging will be manufactured in the future. Intel is proud to be a partner.” Following the announcement, Intel’s shares increased by approximately 4%, closing at $52.91. This market response underscores the significance of the deal for Intel’s foundry aspirations: in the most recent fiscal year, Intel’s Foundry division generated just $307 million in revenue from external customers, a number that pales in comparison to Taiwan Semiconductor Manufacturing Company (TSMC), which earns tens of billions from external clients annually. Should Terafab be even partially realized, it could completely transform Intel Foundry’s commercial landscape.

      Intel’s recovery and the requirements for this venture

      Tan assumed leadership of an Intel that was in dire straits. The company had fallen behind TSMC and AMD across most major product categories, its manufacturing roadmap faced multiple delays, and its foundry business, aimed at producing chips for external clients akin to TSMC, had garnered limited interest outside of government-backed contracts under the US CHIPS and Science Act. Tan undertook aggressive restructuring measures: thousands of layoffs, a tighter focus on the 18A and 14A process nodes as the basis for the foundry proposition, and a conscious effort to position Intel’s domestic manufacturing capability as a