Tesla Optimus 3 Mass Production Accelerates as Taiwanese Suppliers Begin Stocking Components
nashnova research
Two Taiwanese suppliers have started delivering core components for Tesla's Optimus 3 humanoid robot, with mass-production timelines pointing to H2 2025 through 2027 — a real-world signal that Musk's robot roadmap is moving from slides to factory floors.
What exactly is the supply chain shipping?
Mirle Automation has begun supplying harmonic reducers and joint modules — precision parts that let a humanoid robot's joints move accurately. Think of them as the robot's "skeletal joints."
Asia Optical is providing spherical and aspherical lenses for Optimus 3's vision system — the optics that let the robot see.
This means → the supply chain now covers motion + vision, the two core subsystems. This is system-level procurement, not a single-part trial order.
Does the production timeline add up?
Musk said in March: low-volume production starts this summer; high-volume begins in 2027.
Both Taiwanese suppliers' shipping schedules — ramping from H2 2025 through 2027 — align closely with that timeline.
In plain terms = suppliers are committing real capital to inventory. That signals firm orders, not a bet on a conference-slide date.
How big are the capacity targets?
Tesla is building its first robot production line in Fremont, targeting one million units per year.
A second line will go into the Texas Gigafactory, with a planned annual capacity of ten million units.
This means → Tesla is planning the robot as its next "EV-scale" mass-market product, not a lab experiment.
Why does the Thailand joint venture matter?
Mirle and Shenzhen-based Kedali Industry have set up a joint-venture factory in Rayong, Thailand, producing harmonic reducers, actuators, and other precision parts.
Key detail: the plant runs on Kedali's existing facilities. This means → no greenfield build is needed, so production can start much faster.
This reflects a supply-chain footprint expanding beyond Taiwan into Southeast Asia — the logic is cost reduction and capacity redundancy.
What still stands between stockpiling and real mass production?
Supplier-level stockpiling validates component-level feasibility, but full-unit assembly, software integration, and yield ramp remain open questions.
In plain terms = parts on a production line ≠ robots walking out of a factory. Between small-batch and one-million-unit scale lies a long manufacturing gauntlet.
Whether Tesla can cross that gap on schedule will be the defining test of how deep its robot strategy truly goes.
市场有风险，内容仅供研究参考，不构成投资建议。