Elon Musk confirms Tesla and SpaceX AI will keep ordering Nvidia chips despite AI5 development

Tesla AI Chip Development Continues Alongside Nvidia Partnership

Elon Musk confirmed late Wednesday that both Tesla and SpaceX AI expect to continue ordering Nvidia chips at scale, even as the electric vehicle maker advances its own chip development efforts 1

1

. This marks the first public reference to the combined entity following SpaceX's acquisition of xAI in an all-stock deal last month, ahead of a potential blockbuster initial public offering for SpaceX later this year 1

1

. The announcement signals Tesla's dual-track strategy of maintaining partnerships with established chip suppliers while building proprietary artificial intelligence chips to power its autonomous ambitions.

Tesla's AI5 Self-Driving Chip Promises Efficiency Gains

Tesla is designing its fifth-generation AI chip, which Musk claims will "punch far above its weight" due to the company's efforts to "make maximally effective use of every circuit" 2

2

. The AI5 chip is primarily optimized for AI edge compute in the Optimus humanoid robot and Robotaxi, though it can also be used for training in data centers 1

1

. These chips power Tesla's autonomous driving systems, including the Full Self-Driving software, which Musk said expects a wide release update in a few weeks 1

1

. The dual-purpose design reflects Tesla's broader vision of deploying AI across multiple product lines, from vehicles to humanoid robots.

Terafab Project and Future Chip Roadmap

Musk announced last Saturday that Tesla's Terafab project to manufacture artificial intelligence chips will launch in seven days, focusing on producing the AI5 chip as a crucial element of Tesla's self-driving efforts 1

1

2

2

. Looking ahead, Musk revealed that a single AI6 chip could match a dual SoC AI5 within "the same half reticle and same process node," suggesting significant performance improvements in the next generation 2

2

. This roadmap indicates Tesla's commitment to rapid iteration in chip design, potentially reducing costs and power consumption while increasing computational capabilities for both Optimus robots and Robotaxis.

Maintaining Strong Nvidia Relationship Despite In-House Development

In the same thread, Musk shared that he was a "huge admirer" of both Nvidia and Jensen Huang, praising Nvidia's status as the leading company by market capitalization in the world 2

2

. "That market cap is well-deserved," Musk said, adding that both SpaceX and Tesla "expect to continue ordering Nvidia chips at scale" 2

2

. This suggests that Tesla's in-house chip development complements rather than replaces Nvidia's offerings, likely with Tesla chips handling edge compute tasks while Nvidia hardware powers data center applications for training Full Self-Driving models. The strategy allows Tesla to optimize for specific use cases while leveraging Nvidia's proven infrastructure for computationally intensive tasks, positioning both companies to benefit from the growing demand for AI hardware across the automotive and robotics sectors.