Elon Musk on Thursday signaled that Tesla Inc. may need to build a “gigantic chip fab” to support its artificial intelligence ambitions, and floated the possibility of working with struggling U.S. chipmaker Intel Corp.
Speaking at Tesla’s annual shareholder meeting, Musk outlined plans for the company’s fifth-generation AI chip, which will power its autonomous driving technology, including the Full Self-Driving software. “You know, maybe we’ll do something with Intel,” Musk said to an enthusiastic audience. “We haven’t signed any deal, but it’s probably worth having discussions with Intel.”
Intel, which operates its own manufacturing facilities, has fallen behind Nvidia Corp. in the fast-growing AI chip market. The U.S. government recently took a 10% stake in Intel as the company seeks external customers for its most advanced manufacturing technology. Following Musk’s remarks, Intel shares rose 4% in after-hours trading. Intel declined to comment on potential talks with Tesla.
Musk also celebrated a victory at the meeting as shareholders approved a $1 trillion pay package tied to the company’s performance over the next decade, cementing his influence and supporting his vision of transforming Tesla into a leader in AI and robotics.
Tesla has previously indicated that it is working with Taiwan Semiconductor Manufacturing Co. (TSMC) and Samsung Electronics Co. on AI chips. The AI5 chip, which will succeed the company’s current fourth-generation chip, is expected to see limited production in 2026, with high-volume manufacturing likely in 2027. Musk noted that AI6, the next iteration, would use the same facilities but deliver roughly double the performance, with volume production targeted for mid-2028.
“Even when we extrapolate the best-case scenario for chip production from our suppliers, it’s still not enough,” Musk said. “So I think we may have to do a Tesla terafab. It’s like giga but way bigger. I can’t see any other way to get to the volume of chips that we’re looking for.” He described the fab as potentially handling at least 100,000 wafer starts per month, a measure of semiconductor output.
Musk emphasized that the new chips would be inexpensive, power-efficient, and optimized for Tesla software. He projected the AI5 chip would consume roughly a third of the power used by Nvidia’s Blackwell chip while costing only 10% as much to produce.
“I’m super hardcore on chips right now, as you may be able to tell,” Musk said. “I have chips on the brain.”
Industry analysts noted that Tesla’s move reflects the company’s broader push to vertically integrate its AI hardware, a strategy designed to reduce reliance on third-party suppliers and control the pace of innovation. By potentially building its own massive fabrication facility, Tesla would join a small group of companies capable of high-volume semiconductor manufacturing—a costly and complex undertaking that has challenged even established chipmakers.
Tesla did not provide further details on the potential location, timeline, or funding for the proposed fab. Musk’s comments, however, underscore the company’s growing focus on AI hardware as central to its future growth, particularly as competition intensifies in autonomous driving and AI-driven robotics.
