The GPU Gold Rush: What Larry Ellison and Elon Musk Begged Nvidia’s CEO for at Nobu

In a recent high-profile dinner at Nobu in Palo Alto, Larry Ellison, chairman of Oracle, and Elon Musk, founder of Tesla and SpaceX, found themselves in a unique situation while dining with Jensen Huang, CEO of Nvidia. Both tech giants are scrambling for an essential resource: graphics processing units (GPUs), which are crucial for their ambitious AI projects. With demand skyrocketing, the two billionaires shared their urgent appeal for more GPUs, something that even their significant financial resources cannot easily secure.

Ellison’s Oracle is currently at the forefront of building advanced data centers, boasting 162 facilities already operational or under development, with aspirations to reach up to 2,000. The Energized demand for computing resources from AI developers has necessitated a staggering configuration of over 32,000 GPUs in some of Oracle’s largest data centers. However, to stay competitive in the burgeoning AI sector, Oracle is preparing to unveil a jaw-dropping GPU cluster comprising 131,072 units from Nvidia’s latest Blackwell series.

Despite Oracle’s innovations in rapid data transfer, using their own random direct memory access (RDMA) networking technology, they’ve hit a wall due to the limited supply of these powerful chips. In their latest fiscal quarter, Oracle’s cloud computing segment saw a remarkable 45% increase year-over-year, generating $2.2 billion in revenue. Still, supply chain issues have held it back from achieving even greater heights.

On the other hand, Musk’s Tesla is on a parallel path, aiming to launch a significant cluster of 50,000 GPUs this year to enhance its autonomous driving capabilities. Like Oracle, Tesla is not alone in the race; other major tech firms, including Meta, are also consuming GPU resources aggressively. Meta, for example, crafted its Llama 3.1 large language model using 16,000 of Nvidia’s flagship GPUs, planning to expand its capacity to an astonishing 600,000 by year’s end.

This epic GPU chase isn’t merely a race for technology supremacy; it’s a reflection of a broader arms race in the corporate and global landscapes as firms push to lead the charge in AI innovation. Ellison, with tongue in cheek, emphasized the restaurant exchange, humorously exclaiming, “Take our money! We desperately need more of those GPUs!” Their unmistakable plea illustrates the intensity of the competition.

Nvidia, the industry leader in this space, has seen a phenomenal 154% surge in data center revenue compared to last year, raking in $26.3 billion during the second quarter of fiscal year 2025. As companies continue to ramp up investments, with projections indicating another decade of significant spending, the tech landscape is undoubtedly transforming. Ellison estimates that future investments in AI could reach unparalleled levels as firms vie for dominance.

Both Oracle and Tesla have committed substantial resources, with Oracle planning to double its capital expenditures to $13.8 billion in fiscal 2025 and Tesla aiming to spend over $10 billion this year. Comparatively, tech titans like Microsoft and Amazon lead the pack in capital expenditures, with Microsoft earmarking $55.7 billion and Amazon potentially reaching up to $60 billion in 2024 for AI infrastructure.

As the demand for GPUs continues to grow exponentially, investors are keenly observing how these developments could shape the future of the AI market. For those looking to dive into technology stocks, Nvidia remains a prominent name, but it’s crucial to remain informed about the fluid nature of this sector and the competitive strategies employed by major players.

With growth trajectories pointing upwards, understanding the interplay between demand for tech components and financial strategies will be critical for anyone seeking an advantage in the evolving technology landscape. As major companies position themselves in what could be described as the modern equivalent of an energy race, being informed about the latest trends in artificial intelligence and data processing may yield lucrative opportunities for savvy investors and tech enthusiasts alike.