⏳ In Brief
-
Tesla has disbanded the Dojo supercomputer team, ending internal AI compute development
-
Peter Bannon, lead chip architect, has exited the company
-
Around 20 Dojo team members have joined a new startup, DensityAI
-
Remaining employees are reassigned to other Tesla engineering projects
-
Tesla will now rely on Nvidia, AMD, and Samsung for its compute infrastructure
Tesla Shuts Down Dojo Team, Resets AI Compute Ambitions
Tesla has officially shut down its Dojo supercomputer team, marking a significant shift in how the company plans to power its AI systems going forward.
The Dojo initiative was originally designed to support autonomous vehicle training by processing video data from millions of Tesla cars. The move away from internal infrastructure reflects a clear pivot toward external compute partnerships.
The project’s termination aligns with a broader realignment of Tesla’s AI strategy, prioritizing speed and scale over proprietary hardware development.
“Tesla has disbanded its Dojo team and reassigned engineers to other hardware and data center roles,” a source familiar with the matter said.
Leadership Exit and Team Dispersal
Peter Bannon, a former Apple chip designer and the executive behind Dojo, has now left Tesla. Around 20 engineers from the project have moved to a new AI venture, DensityAI.
The rest of the Dojo team is being redistributed across Tesla’s internal hardware, compute, and data engineering divisions, including those supporting Full Self-Driving (FSD) infrastructure.
While the company has not issued a formal statement, internal reassignment and leadership exit confirm that Tesla is no longer pursuing in-house AI supercomputing.
Why Tesla Pulled the Plug
The decision appears driven by multiple factors, including cost, timeline delays, and the rapid pace of development from industry leaders like Nvidia.
Instead of continuing to invest in custom Dojo chips and compute clusters, Tesla will now lean on external providers. These include Nvidia for AI GPUs, Samsung for chip fabrication, and AMD for performance processing.
This allows Tesla to tap into cutting-edge silicon without carrying the full burden of long-term hardware development a challenge even large-scale tech companies struggle with.
A New Phase in Tesla’s AI Strategy
By disbanding the Dojo team, Tesla effectively signals its intent to outsource compute power, while keeping AI model development and data pipelines in-house.
The company continues to scale its FSD beta testing, AI labeling systems, and software stacks, but now without the additional risk and capital required to maintain a proprietary training system.
“This is about focusing Tesla’s energy on product-side AI, not compute-side infrastructure,” noted a senior engineer familiar with the shift.
What Happens to Tesla’s AI Momentum Now
Despite the internal change, Tesla’s AI ambitions remain aggressive. The company still plans to build robotaxi platforms, integrate real-time AI into its Energy division, and deploy AI-enhanced robotics in factories.
The change means Tesla must now ensure deep collaboration with partners to handle workloads at scale, particularly as it collects more driving data and pushes FSD updates globally.
While Dojo is gone, Tesla’s AI roadmap is still active but now fueled by outsourced compute power.
Conclusion
Tesla’s decision to disband its Dojo supercomputer team marks the end of its ambitious effort to control the full AI training stack internally.
With Peter Bannon’s exit, the shift toward Nvidia, AMD, and Samsung partnerships suggests Tesla now values agility over autonomy in its AI infrastructure. Whether this trade-off accelerates or slows its AI goals remains to be seen, but it’s a defining change in strategy.
📈 Trending Right Now
- OpenAI rolls out GPT‑5 with free access and major upgrades
- GPT-5 outperforms GPT-4 with deep reasoning and multimodal capabilities
- What you can actually do with GPT-5 right now
- Here’s how to use GPT‑5 for free today
- DICE confirms Battlefield 6 to use AI bots when lobbies aren’t full
For more AI stories, visit AI News on our site.
