On March 21, Elon Musk announced the Terafab project as a massive, all-in-one semiconductor facility, designed to produce huge amounts of AI computing power, and aims to produce up to 1 terawatt compute per year. The factory plans to start with 100,000 wafers per month, with a pilot phase costing between $20 to 25 billion.
How does Intel play a part in the Terafab project?
Intel will be joining the project as a manufacturing partner, bringing its chip-building expertise to Elon Musk's ambitions. The company is expected to accelerate Terafab’s goals with its end-to-end capabilities spanning the full semiconductor stack.
Intel's 18A process node, which is already running at its Chandler, Arizona facility at approximately 40,000 wafer starts per month, will likely become the foundation for Terafab's initial production.
For Intel CEO Lip-Bu Tan, the partnership comes as a critical move to bring Intel back into a leading position in the AI chip manufacturing business. The company is facing tough competition from rivals like Nvidia and AMD, which have overtaken the industry by focusing on design and outsourcing manufacturing.
Now, after a visit from Elon Musk, Tan, in a separate X post, described Terafab as a “step change,” meaning it could significantly transform the industry. In addition, Intel will also get backing from the most influential global companies, including SpaceX, xAI and Tesla.
The partnership could also help Intel strengthen its IDM 2.0 strategy, which plans to design and manufacture chips at scale. In addition, the partnership positions Intel as a key player powering Terafab, which Musk sees as the most ambitious chip-building project ever.
Post the partnership announcement, Intel’s shares jumped more than 2% as of the latest data.