Tesla Inc (NASDAQ:TSLA) has disbanded its Dojo supercomputer team, with the team leader Peter Bannon departing the company, according to a Bloomberg report.
The remaining Dojo team members are reportedly being reassigned to other data center and computing projects within Tesla.
This marks a significant shift away from Tesla’s years-long effort to develop an in-house AI training supercomputer and custom chips for its autonomous vehicle technology.
The Dojo project, launched in 2021, was designed to train machine-learning models rapidly using data from Tesla vehicles, powering features like Autopilot, Full Self-Driving, and Tesla’s humanoid robot Optimus.
Tesla developed custom chips like the D1 and was working on the D2, but the company now plans to rely more on external technology partners such as Nvidia, AMD, and Samsung for computing and chip manufacturing, per the report.
Following Bloomberg’s report, Tesla CEO Elon Musk confirmed Tesla has shifted its chip strategy.
"It doesn't make sense for Tesla to divide its resources and scale two quite different AI chip designs," Musk wrote in a post on his social media platform X on Thursday evening.
"The Tesla AI5, AI6 and subsequent chips will be excellent for inference and at least pretty good for training. All effort is focused on that.”
Wedbush noted that Tesla appeared to be giving up on its plans to develop a custom solution for AI compute.
“We see this news as a reminder that developing custom ASICs is not easy and that while Amazon and Google have clearly had success with internal silicon efforts, that these products have only matured and yielded benefits over an extended timeframe,” the analysts wrote.
Shares of Tesla edged 0.4% lower before Friday’s opening bell, down more than 20% so far in 2025.