On April 23, 2026, Elon Musk, CEO of Tesla, announced that both Tesla and SpaceX will adopt Intel's 14A manufacturing process for the ambitious Terafab project. This declaration was made during Tesla's Q1 2026 earnings call. The Terafab project, initially announced in March 2026 and slated for Austin, Texas, is envisioned as a large-scale advanced AI chip complex.
Musk stated that the selection of the 14A process was based on its anticipated maturity by the time Terafab scales up. He remarked, "Given that by the time Terafab scales up, 14A will be probably fairly mature or ready for prime time. 14A seems like the right move." This partnership marks Tesla as the first major external customer for Intel's 14A technology, a development that is expected to significantly aid Intel's efforts to expand its foundry business.
The Terafab project is designed to manufacture processors crucial for autonomous vehicles, humanoid robotics, and data center systems, including those deployed in space. Musk has outlined plans for two distinct fabrication facilities within the Terafab complex. One facility is intended to support chips for vehicles and robotics, while the second will focus on infrastructure for space-based data centers.
According to the announced division of responsibilities, Tesla will handle the establishment of a research and development facility at its Texas Gigafactory campus. This pilot line, with an estimated investment of approximately $3 billion, will have a monthly capacity of a few thousand wafers and will primarily serve for technical validation and experimentation with new semiconductor manufacturing concepts. SpaceX will be responsible for the high-volume manufacturing aspect, building and operating the high-capacity chip production facilities.
Intel's 14A process is a 1.4nm-class advanced process node. While not yet fully developed, it is projected to offer a 15% to 20% performance increase over the 18A node, a 30% increase in density, and a 25% to 35% reduction in power consumption. The Terafab project aims for an ultimate output of one terawatt of computing power annually, a substantial increase compared to the estimated half-terawatt currently produced in the United States. Achieving this level of compute will necessitate large-scale semiconductor fabrication, advanced chip packaging, and extensive deployment in data center infrastructure.