On April 7, 2026 (local time), Intel confirmed its involvement in the Terafab project—a collaborative effort spearheaded by Tesla, SpaceX, and xAI. The initiative seeks to revolutionize silicon wafer manufacturing technology, aiming for an annual computational output of 1 terawatt (1 trillion watts) to accelerate advancements in AI and robotics. Intel emphasized that its expertise in chip design, production, and packaging will be instrumental in helping Terafab achieve this ambitious target.

Based in Austin, Texas, the Terafab project will utilize a cutting-edge 2-nanometer manufacturing process, integrating logic chips, memory modules, and advanced packaging into a unified production chain. The facility aims to produce 100 to 200 billion chips annually. These chips will fall into two categories: one tailored for Tesla’s autonomous driving systems and Optimus humanoid robots, and the other optimized for space environments to support SpaceX’s orbital data centers.

Musk revealed that to overcome limitations in terrestrial power supply, Terafab intends to shift 80% of its computational capacity to space. This will involve deploying a global network of micro AI satellites capable of performing AI computations and data processing directly in orbit.

Intel’s participation brings vital industrial execution capabilities to the Terafab project. This partnership is expected to propel chip manufacturing technology toward unprecedented performance levels and expanded applications.