devxlogo

Elon Musk diverts Nvidia AI chips from Tesla

Elon Musk diverts Nvidia AI chips from Tesla

Musk Nvidia

Elon Musk has admitted to diverting Nvidia’s latest AI chips, originally intended for Tesla, to his other companies, including his AI startup xAI and social media platform Twitter. In a statement on Tuesday, Musk explained that the chips, used for training neural networks, would have otherwise “just sat in a warehouse.

Tesla is known for its electric vehicles, but Musk has been rebranding the company as an AI and robotics-focused entity. He has committed to improving Tesla’s AI training and inference capabilities.

In April, Musk announced plans to acquire chips equivalent to 85,000 Nvidia H100s by the end of the year for AI training purposes. At the end of the first quarter, Tesla had around 35,000 of these chips. The timing of Musk’s disclosure could be problematic, as he awaits a crucial June 13 court decision regarding his substantial pay package from 2018.

Tesla shareholders may be concerned that the diversion of AI chips to Musk’s other ventures could negatively impact Tesla. These concerns have grown since Musk’s xAI startup began using resources, including those from Tesla.

Musk reallocates Nvidia chips

An expansion of Tesla’s Texas site to house a new data center is nearing completion. This facility will accommodate 50,000 H100 chips for Full Self-Driving (FSD) training, addressing concerns about potential delays in Nvidia chip shipments to Tesla. Musk emphasized the complexity of integrating 50,000 H100 chips into a coherent system, a feat no company has yet achieved.

It is unclear why Tesla initially ordered the chips without having a functioning data center to use them. One possible explanation is the high demand for Nvidia’s advanced H100 training chips, which made Musk reluctant to forgo an available delivery slot. Tesla has reportedly invested $1 billion in the first quarter to more than double its computing capacity compared to the previous three months.

See also  Internet divides remote Marubo tribe deeply

While it is not clear if diverting some of the chips harmed Tesla materially, Musk’s actions highlight the challenges of managing a publicly traded company alongside several private ventures that compete for the same resources. Musk has a history of reallocating resources as he deems necessary; during a trial regarding his pay package, the CEO confirmed that he had dispatched 50 Tesla engineers to assist at Twitter. This scenario demonstrates the complexity and potential risks involved when a business leader manages multiple companies with overlapping needs and resources.

devxblackblue

About Our Editorial Process

At DevX, we’re dedicated to tech entrepreneurship. Our team closely follows industry shifts, new products, AI breakthroughs, technology trends, and funding announcements. Articles undergo thorough editing to ensure accuracy and clarity, reflecting DevX’s style and supporting entrepreneurs in the tech sphere.

See our full editorial policy.

About Our Journalist