Nvidia was instructed by Elon Musk to deliver thousands of AI chips reserved for Tesla to X and xAI.

• Nvidia emails obtained by CNBC reveal that Elon Musk instructed the chipmaker to prioritize shipments of processors to X and xAI over Tesla.
• Tesla, according to Musk, has the potential to become a significant player in the field of artificial intelligence, with the company heavily investing in Nvidia's AI processors.
• Nvidia was delayed in receiving over $500 million in processors by months because Musk instructed them to prioritize Tesla over Nvidia.

Elon Musk asserts that he intends to become a leader in AI and robotics, which he believes will necessitate the acquisition of expensive processors from Tesla.

Tesla plans to increase the number of active H100s, Nvidia's top AI chip, from 35,000 to 85,000 by the end of this year. Additionally, the company will spend $10 billion this year on combined training and inference AI.

Nvidia senior staff's emails suggest that Musk exaggerated Tesla's procurement to shareholders and diverted a significant shipment of AI processors reserved for Tesla to his social media company X.

Shares slipped as much as 1% on the news in premarket trading.

Musk delayed Tesla's receipt of more than $500 million in GPUs from Nvidia by months, likely adding to delays in setting up the supercomputers Tesla needs for autonomous vehicles and humanoid robots.

"According to an Nvidia memo from December, Elon has decided to prioritize the deployment of an H100 GPU cluster at X instead of Tesla by redirecting 12k of shipped H100 GPUs originally intended for Tesla to X. In return, original X orders of 12k H100 GPUs slated for January and June will be redirected to Tesla."

Tesla's ongoing layoffs could cause further delays with an "H100 project" at its Texas Gigafactory, according to a recent Nvidia email from late April. The email also noted that Elon Musk's comment on the first-quarter Tesla call "conflicts with bookings," and his April post on X about $10 billion in AI spending also "conflicts with bookings and FY 2025 forecasts."

Tesla shareholders are becoming increasingly frustrated with CEO Elon Musk, who is facing criticism for not fulfilling his obligations to the company while also managing multiple other businesses that require his time, capital, and resources.

Nvidia, X, and Tesla did not respond to requests for comment.

Musk is a CEO of multiple companies, including Tesla, SpaceX, Neuralink, The Boring Co., and X, which he acquired for $44 billion in late 2022. He is also the founder of brain-computer interface startup Neuralink and launched his AI startup, xAI, in 2023.

xAI and X are closely linked. In a November post, Musk stated that X Corp investors would own 25% of xAI. Furthermore, xAI utilizes some of its training and inference for the large language models behind its chatbot, Grok, in X data centers.

Grok, originally known as Truth GPT, is a politically incorrect chatbot with a rebellious streak, as pitched by Musk, who sees it as a competitor to OpenAI's ChatGPT and other generative AI services.

Despite Musk's numerous business endeavors, Tesla's sales are declining due to an aging electric vehicle lineup and increased competition. Additionally, the company's reputation in the U.S. has been negatively impacted by Musk's "antics" and "political rants," according to the Axios Harris Poll 100 survey.

Tesla's stock price is down 29% this year.

Musk has been urging investors to concentrate on future products that he has been promising for years but has not yet delivered, such as AI software to transform existing cars into self-driving vehicles, dedicated robotaxis that can generate income for their owners, and a driverless transportation network.

""If someone doubts Tesla's ability to solve autonomy, they should not invest in the company, as we will and are capable of doing so," Musk stated during the April earnings call."

Tesla needs Nvidia's GPUs, which are designed for AI training and workloads, to reach its destination. However, these chips are in short supply due to high demand from various industries, including gaming, autonomous vehicles, and cloud computing.

'Consuming every GPU that's out there'

Nvidia, currently the third-most-valuable company globally with a $2.8 trillion market cap, is struggling to meet demand due to high consumption of GPUs by cloud service providers and AI model developers, as stated by Nvidia CEO Jensen Huang during a May earnings call following the chipmaker's third consecutive quarter of over 200% revenue growth.

During a February earnings call, Huang stated that Nvidia strives to fairly allocate resources and avoid unnecessary allocations, questioning the need for allocation when data centers are not yet ready.

Huang mentioned xAI as a customer of Nvidia's next-generation Blackwell platform during a May call with six of the largest tech companies globally, including Tesla.

Musk likes to tout his infrastructure spending at both companies.

Tesla has pledged to construct a $500 million "Dojo" supercomputer in Buffalo, New York, and a "super dense, water-cooled supercomputer cluster" at its factory in Austin, Texas. This technology could aid Tesla in developing the computer vision and LLMs required for robots and self-driving vehicles.

In addition to competing with OpenAI, Anthropic, Google, and others in developing generative AI products, Musk is also working to create "the world's largest GPU cluster" in North Dakota, with some capacity available online in June, as stated in an internal Nvidia email from February.

The memo outlined a "Musk mandate" to ensure that all 100,000 chips were made available to xAI by the end of 2024. The document also mentioned that the LLM behind xAI's Grok was relying on Amazon and cloud infrastructure, with X providing additional data center capacity.

The Information previously reported some details of xAI's data center ambitions.

On May 26, xAI announced the successful completion of a $6 billion financing round, led by the same investors who backed Musk's Twitter acquisition. The company was established in March 2023, but Tesla kept its formation a secret until Musk publicly unveiled the startup four months later.

Conflicts of interest

In January, Musk stated on X that he wants more control over Tesla before pushing further in the direction of AI, despite his previous claims that the company is a leader in this field.

He stated that he was uncomfortable with Tesla becoming a leader in AI and robotics without having at least 25% voting control, as it would be influential but not so much that he could be overturned.

Tesla's recent proxy filing shows that Musk owns 20.5% of the company's outstanding shares, which includes options granted to him as part of his unique 2018 CEO compensation package. A Delaware court has ruled that this compensation be rescinded. The post-trial proceedings are ongoing and may be subject to appeal.

In the January post, Musk stated that if he cannot achieve his desired ownership mark, he would prefer to create products outside of Tesla. He is already doing that at xAI.

Some longstanding bulls, including the company's largest retail shareholder, Leo Koguan, and Gerber Kawasaki's Ross Gerber, were offended by Musk's comments at the time, which they characterized as "blackmail."

Musk's decision to let his private companies procure critical hardware before Tesla exposes his conflicts of interest.

"Fleming stated that when a person like Mr. Musk serves as a fiduciary to multiple companies, the law recognizes this creates a conflict. If a fiduciary owes duties to two or more competing companies, they may unintentionally divert corporate opportunities from one company to another."

Fleming, who often represents public company investors in shareholder disputes, stated that in such situations, other executives would be the best decision-makers, while those with conflicts should abstain.

Fleming stated that Mr. Musk has not typically followed that path.

Musk hasn't been shy about intermingling corporate resources among his companies.

Musk hired numerous Autopilot software engineers and other technical and administrative staff from Tesla to assist him in making significant changes at Twitter. Some employees even work for both Musk companies simultaneously.

Musk has attracted employees away from Tesla to xAI, including machine-learning scientist Ethan Knight and at least four other former Tesla employees involved in Autopilot and big data projects.

A former Tesla supply chain analyst, who requested anonymity to discuss confidential information, revealed to CNBC that Musk views his companies as an extension of his personal brand and has the freedom to make any decisions he desires, including Tesla's 2016 acquisition of SolarCity, where he served as chairman and a significant shareholder.

The decision to redirect a large shipment of chips from Tesla to X is extreme, given the scarcity of Nvidia's technology. This means that Tesla willingly gave up valuable time that could have been used to build out its supercomputer cluster in Texas or New York and advance the models behind its self-driving software and robotics.

WATCH: Nvidia is a one of one company