Billionaire Elon Musk said on March 18 that his companies expect to continue purchasing artificial intelligence chips from Nvidia at large scale, underscoring the ongoing demand for high-performance computing hardware across his expanding technology ventures. The comments come as Musk’s companies deepen their focus on AI development spanning autonomous vehicles, robotics, and data infrastructure.
Musk, who is widely known as the world’s richest individual with a reported net worth of approximately $832.1 billion according to Forbes, stated, “I am a huge admirer of Nvidia and Jensen btw. That market cap is well-deserved.” He added, “SpaceX AI and Tesla expect to continue ordering Nvidia chips at scale,” signaling continued reliance on Nvidia’s hardware even as his companies develop their own in-house semiconductor technologies.
I am a huge admirer of Nvidia and Jensen btw. That market cap is well-deserved.
— Elon Musk (@elonmusk) March 19, 2026
SpaceX AI and Tesla expect to continue ordering Nvidia chips at scale.
The remarks were tied to a broader discussion of Tesla’s next-generation AI chip, known as AI5, which is being developed to support the company’s autonomous driving ambitions. Tesla is currently working on its fifth-generation chip to power systems such as its Full Self-Driving software, which remains central to the company’s long-term strategy around automation and mobility.
Describing the chip’s capabilities, Musk said, “AI5 will punch far above its weight, because the entire Tesla AI software stack is designed to make maximally effective use of every circuit. We co-signed our AI software and hardware.” He emphasized that the chip is engineered with efficiency in mind, integrating both software and hardware design to maximize performance.
Musk further explained the intended use of the AI5 chip, noting that while it can support data center training workloads, its primary purpose lies in edge computing applications. “Bear in mind that AI5, while it can be used for training in data centers, is primarily optimized for AI edge compute in Optimus and Robotaxi,” he said, referring to Tesla’s humanoid robot project and its planned autonomous ride-hailing platform.
Tesla’s forthcoming AI5 uses a half-reticle design, which is crucial for yield. A reticle defines the imaging area of a lithography machine, fitting two chips per shot effectively doubles yield.
— phil beisel (@pbeisel) March 18, 2026
This means the Tesla chip design team had to carefully manage die features, for… https://t.co/vLWkNpPfvS pic.twitter.com/LgFmNJ7dZa
Looking ahead, Musk also pointed to continued improvements in future chip generations. “There is still significant room for improvement. In the same half reticle and same process node, we think a single AI6 chip has the potential to match a dual SoC AI5,” he said, suggesting that subsequent iterations could deliver substantial gains in efficiency and processing power.
In parallel with these hardware developments, Musk indicated that Tesla expects to release a broader update to its Full Self-Driving (Supervised) software in the coming weeks, marking another step in the company’s incremental rollout of autonomous features.
The comments also follow recent corporate developments involving Musk’s aerospace company SpaceX. Last month, SpaceX acquired xAI in an all-stock transaction, a move that has drawn attention ahead of a potential initial public offering for SpaceX later this year. Musk’s reference to “SpaceX AI” in his latest remarks appears to be the first time he has publicly described the combined entity using that name.
Together, the statements highlight the dual-track approach across Musk’s companies, combining continued partnerships with established chipmakers like Nvidia while simultaneously advancing proprietary AI hardware designed to support increasingly complex autonomous and robotic systems.
By
By
