Key Points
- Tesla is shutting down its in-house Dojo supercomputer project.
- The project’s leader, Peter Bannon, is leaving the company, and the team is being disbanded.
- Tesla will now rely more on outside partners like Nvidia, AMD, and Samsung for its AI computing needs.
- Elon Musk confirmed the move, saying it didn’t make sense to pursue two separate chip designs.
Tesla is shutting down its ambitious Dojo supercomputer project, and its leader, Peter Bannon, is leaving the company. The move marks a major strategy shift for the automaker, which will now rely more on outside partners for the powerful chips needed for its self-driving technology.
The Dojo system was a custom-built supercomputer using Tesla’s own in-house D1 chips. Its purpose was to train the massive AI models that power Tesla’s Autopilot, Full Self-Driving features, and its Optimus robot. The project was once seen as a key competitive advantage, with some analysts estimating it could add $500 billion to Tesla’s value.
CEO Elon Musk confirmed the news on the social media platform X, stating that it didn’t make sense to divide the company’s resources between two different AI chip designs.
The Dojo team is now being disbanded. Many of its workers have already left for a new AI startup, DensityAI, which was founded by Dojo’s former head. The remaining employees are being reassigned to other projects within Tesla.
In the future, Tesla plans to increase its reliance on tech giants like Nvidia and AMD for its computing power and will work with Samsung for chip manufacturing. This pivot comes as Tesla faces an exodus of key talent, falling sales, and increased competition in the EV market. It’s a sign that the company is rethinking its “do everything in-house” approach to technology.
