Tesla is accelerating the pace of its artificial intelligence hardware development, with Chief Executive Elon Musk saying the company is aiming for a nine-month design cycle for its custom AI chips, a timeline that would be unusually fast for the semiconductor industry.
Musk said Tesla’s next-generation AI5 chip design is nearly complete, while work on its successor has already begun. “Our AI5 chip design is almost done and AI6 is in early stages, but there will be AI7, AI8, AI9 … aiming for a 9-month design cycle,” Musk wrote on X, also encouraging engineers to join Tesla’s AI hardware team.
See also: Tesla Patent Points to Software Path for Extending Life of Older FSD Hardware
The comments suggest Tesla is attempting to shift from multi-year chip development cycles toward a rapid iteration model more typical of software development. If achieved, the pace would put Tesla among the fastest-moving custom chip designers, ahead of most automotive and AI hardware peers.
Industry analysts note, however, that faster design cycles do not automatically translate into rapid deployment. Even after a chip design is finalized, manufacturing, validation and automotive qualification can take 12 to 24 months or longer before high-volume production and integration into vehicles.
See also: Tesla Starts Production at Texas Lithium Refinery, Showcases New Process
Tesla’s current production vehicles are based on its AI4 hardware platform, which supports the latest versions of Full Self-Driving (Supervised) by enabling real-time neural network inference for perception and planning. The upcoming AI5 chip is expected to offer a significant increase in performance and efficiency, allowing Tesla to run larger end-to-end neural networks and process video data more quickly.
Musk has said AI5 is expected to be manufactured by TSMC and Samsung Electronics. Future generations, including AI6 and beyond, are planned for production at Samsung’s new Texas facility under a $16.5 billion agreement announced earlier this year.
See also: Tesla Posts Record Sales in Japan as Retail Expansion Defies Global Slowdown
The push for faster iteration follows challenges with Tesla’s earlier Hardware 3 platform, introduced in 2019. As Tesla’s autonomy models grew larger and more complex, the company acknowledged that AI3 lacked sufficient computing capacity for its latest Full Self-Driving features, prompting a shift toward newer hardware and software-based mitigation strategies.
