55 articles · Last covered
Tesla says its first production-quality Optimus robots are coming off the Fremont line, while Musk outlines the training, hardware and AI work ahead of customer deliveries.
Tesla’s new shareholder letter urges the re-election of Musk, Denholm and Straubel, citing its AI and manufacturing ambitions ahead of the December 3 meeting.
Musk says Optimus and Cybercab will account for most of Tesla’s value, while a factory observer describes a building-first approach to the Texas Optimus plant.
At X Takeover, Musk described the hand durability and supply-chain challenges behind Optimus, saying magnet export restrictions delayed the program last year.
Tesla cut AI5 memory to 72GB and AI6 to 144GB, Musk says, citing Optimus supply needs and lower costs while keeping bandwidth unchanged.
Optimus renders labeled gen3 first appeared in Tesla’s Android app on August 27, a four-build comparison shows. Higher-resolution versions arrived September 5, but the assets do not confirm final hardware or a working robot interface.
Tesla’s reported supplier audits put Ningbo’s automotive expertise in focus, as Georg Stieler highlights the Yangtze River Delta’s role in humanoid robotics.
A decompilation of Tesla app 4.61.0 identifies Optimus charger IDs and an energy-site entry, suggesting software preparation without confirming a consumer launch.
Reports point to a new round of Optimus supplier inspections in Ningbo, while company representatives confirm ongoing robotics work but stop short of verifying the latest visits.
A new drone update shows steel and concrete work advancing at Tesla’s Texas robot factory, while production timing remains less certain than the construction progress.
Chinese supply chain reports suggest Tesla has placed bulk component orders for roughly 5,000 Optimus units, fueling speculation about the humanoid robot’s path to commercial scale.
With the formal inclusion of Intel, Tesla’s Austin "Terafab" project aims to verticalize semiconductor production to a degree never before seen, targeting 1 terawatt of annual compute to support billions of humanoid robots and a burgeoning space-based AI infrastructure.
