tesla pivots: inference chips now the focus after dojo team closure
Palo Alto, Friday, 8 August 2025.
tesla is making a significant shift in its ai strategy. the company will now prioritize inference chips for real-time ai in vehicles and robotics. this follows reports that ceo elon musk ordered the closure of the dojo supercomputer team. this move towards self-sufficiency could reduce tesla’s reliance on nvidia. samsung electronics is already lined up with a $16.5 billion deal to supply ai6 chips. the next-generation ai5 chips are expected by the end of 2026.
strategic shift in ai development
Tesla’s move to concentrate on inference chips signals a key strategic change [1]. Elon Musk stated that splitting resources between two distinct AI chip designs was inefficient [1][3][4]. This decision affects companies like NVIDIA, potentially reducing Tesla’s need for their high-performance computing chips [GPT]. Tesla aims to deploy AI5 and AI6 chips in self-driving vehicles and Optimus robots [1]. The company’s focus will be on chips adept at real-time operation and some training functions [2].
dojo supercomputer team disbanded
The closure of the Dojo supercomputer team marks a significant turn [1]. This team was created to train autonomous driving software [1]. Approximately 20 members of the Dojo team have already joined DensityAI, a startup focusing on data center chips for robotics and AI applications [1][5]. The remaining employees are being reassigned within Tesla to other data center and computing projects [5]. This shift may reflect challenges in the Dojo project’s progress, including construction delays at the Texas data center [5].
impact on tesla’s partnerships
Tesla is adjusting its supply chain strategy for the Dojo supercomputer [5]. Instead of relying solely on TSMC, Tesla plans to diversify its chip supply [1][5]. Samsung will handle front-end production, while Intel will manage module production packaging [5]. Tesla has secured a $16.5 billion deal with Samsung for AI semiconductor supplies through 2033 [1][2][5]. This deal ensures that Samsung’s Texas factory will produce Tesla’s next-generation AI6 chips [5].
talent exodus and restructuring
Tesla faces significant talent departures, exacerbating restructuring challenges [1][2][5]. Key figures, including the head of the Dojo project, have left the company [5]. These departures coincide with slumping EV sales and consumer resistance, adding pressure on Tesla’s stock [1][5]. The company’s restructuring includes layoffs and executive changes [1][2]. Tesla’s share price has suffered due to increased competition and negative consumer sentiment [1].
analyst perspective
Analysts are closely watching Tesla’s strategic pivot [GPT]. Morgan Stanley previously valued the Dojo supercomputer project at $500 billion in 2023 [1]. The shift towards inference chips suggests a move to optimize resources and accelerate the rollout of self-driving capabilities [2][4]. By focusing on inference chips, Tesla aims to enhance the real-time AI performance of its vehicles and robots [1]. Investors should monitor how these changes impact Tesla’s competitive position and long-term growth prospects [GPT].