Tesla and other Elon Musk companies rapidly acquire Nvidia hardware while Tesla publicly commits to building competing internal supercomputer.

In November 2023, Nvidia CEO Jensen Huang faced a question at an all-hands meeting that carried significant business and ethical implications: would the chipmaker follow Apple and Disney in suspending advertising on X due to rising antisemitism and hate speech? This came after X owner Elon Musk had agreed with a post accusing "Jewish communities" of pushing "hatred against whites." Huang's response was diplomatic: Nvidia hadn't advertised on X in a very long time and had no plans to do so. He also emphasized that Nvidia would never make public statements against another business. What Huang omitted was any discussion of Nvidia's deepening commercial ties with Musk's expanding empire.

Nvidia is experiencing unprecedented demand for its graphics processing units and related hardware, fueling a market cap exceeding $2 trillion. Revenue in the latest quarter jumped 265% to $22.1 billion, and last year Nvidia surpassed Intel in total sales. This boom reflects soaring demand for computing power across generative AI workloads, robotics, research, and data center projects. Meanwhile, Musk has made ambitious promises for his companies' AI capabilities—promising developments that require substantial purchases of Nvidia's technology across his portfolio of ventures including xAI, launched formally in July 2023, as well as Tesla and its autonomous vehicle and humanoid robot projects.

The depth of this business relationship was on full display at Nvidia's annual GTC conference in San Jose, California, which drew roughly 16,000 attendees including celebrities like Ashton Kutcher and Kendrick Lamar. Two sessions featured leaders of xAI, with Christian Szegedy, a co-founder and research scientist at xAI and former Google researcher, speaking at a fireside chat, and Igor Babuschkin, another xAI co-founder and OpenAI and Google veteran, presenting on how the startup is using Nvidia GPUs to "accelerate training and inference of their Grok model." In Nvidia's announcement of its Blackwell AI chips, Musk was quoted saying, "There is currently nothing better than NVIDIA hardware for AI." Tesla, which Musk described in January as "an AI/robotics company that appears to many to be a car company," has also made massive Nvidia commitments. Last August, former Tesla AI engineer Tim Zaman posted that a Tesla AI cluster using 10,000 of Nvidia's H100 chips was ready to go live. In January, Musk stated that while a Dojo supercomputer cost $500 million to build, "Tesla will spend more than that on Nvidia hardware this year," and that "the table stakes for being competitive in AI are at least several billion dollars per year at this point."

Yet the Musk-Huang relationship has not always appeared cordial. Last June, Musk called Nvidia "monopolistic" in response to accusations that Nvidia was "spiking the price" of its GPUs due to supply shortages, asserting that "Nvidia will not have a monopoly on large-scale training & inference forever." Despite these tensions, Huang has credited Musk with inspiring Nvidia's foundational AI supercomputer work. At the New York Times' DealBook summit, Huang recounted how Nvidia developed its first AI supercomputer, the DGX system, beginning in 2012, taking approximately five years to perfect before he personally delivered the first unit to Musk for use by OpenAI. Oracle founder Larry Ellison, a close friend of Musk's and former Tesla board member, revealed in December that xAI had secured Nvidia GPUs through Oracle to create the first version of Grok, though Oracle couldn't meet Musk's appetite: "Boy, do they want a lot more GPUs than we gave them."