According to Sawyer Merritt on X, Elon Musk said Tesla may be able to tape out its upcoming AI6 chip in December, noting the schedule could be accelerated using AI in the design process, as shown in Musk’s post on X (according to Elon Musk’s X post). As reported by Merritt, tape-out marks finalization of the chip design before fabrication, implying Tesla is nearing a major milestone for its in-house AI silicon roadmap aimed at autonomy and training efficiency. According to Musk’s X post, the AI6 timeline suggests Tesla is pushing vertical integration to reduce reliance on external accelerators and improve performance per watt for Full Self-Driving training and inference, which could lower cost of compute and expand capacity for model iteration. For suppliers and partners, according to Merritt’s report, a December tape-out would position 2026–2027 for potential bring-up, validation, and early production, creating opportunities in EDA tooling, IP blocks, packaging, and advanced nodes, while signaling competitive pressure for NVIDIA-dependent fleets.

According to Sawyer Merritt, Tesla has officially confirmed the production timelines for its next-generation in-house AI inference chips, with AI5 set to enter production in 2027 and AI6 following in 2028. As reported by Sawyer Merritt, these custom-designed chips are being developed to advance Tesla's autonomous vehicle capabilities and enhance inference performance. The announcement highlights Tesla's commitment to vertical integration and in-house chip innovation, positioning the company to compete in the rapidly evolving autonomous driving market. This strategic move may open new business opportunities as Tesla aims for greater control over AI hardware and improved vehicle autonomy.