Oracle founder Larry Ellison just delivered fantastic news for Nvidia stock investors

Larry Ellison owns 42% of Oracle (ORCL -0.09%)a $465 billion tech giant that is building some of the most powerful data centers for artificial intelligence (AI) development.

Nvidia (NVDA 3.96%) supplies Oracle and most other technology companies with data center chips called graphics processing units (GPUs). Nvidia has seen spectacular growth in its revenue over the past year, and demand for GPUs continues to outstrip supply. However, some investors have begun to question how much longer Oracle and its peers can throw billions of dollars at the chip giant to fuel its AI aspirations.

Concerns that the AI ​​train may be starting to lose steam are a key reason why Nvidia shares are trading 14.5% off their all-time high. But the market may have missed comments this month from Ellison at Oracle’s financial analyst meeting that hinted at more fantastic news for Nvidia investors.

Oracle isn’t close to meeting its AI infrastructure goals

Oracle data centers are unique because they are automated. Each is operationally identical regardless of its size, and because it requires no human workers, it allows the company to build them quickly. In addition, Oracle’s GPU RDMA (Direct Random Memory Access) networking technology enables data to be transmitted point-to-point faster than traditional Ethernet networks.

Because most AI developers pay for compute capacity by the minute, Oracle data centers can offer significant cost savings compared to competing infrastructure. That’s why demand is growing from AI start-ups like OpenAI, Cohere and xAI. Oracle had 85 data centers in operation and another 77 under construction in the first quarter of fiscal 2025 (ended Aug. 31), but Ellison believes it could operate as many as 2,000 in the long term.

Next year, Oracle plans to offer a cluster of 131,072 GPUs, which is a big step up from its largest clusters now, at around 32,000 GPUs. But there’s another difference: The new cluster will use Nvidia’s latest Blackwell chips, which can perform AI inference at a rate 30 times faster than the H100 that Oracle currently uses. In theory, it will allow developers to build the largest AI models in history.

This will bring significant benefits to Nvidia. It generated $26.3 billion in data center revenue in the second fiscal quarter of 2025 (ended July 28), primarily from GPU sales, which was a 154% increase from the year-ago period. The growth rate has slowed compared to previous quarters because the numbers have gotten so big, but Nvidia’s customers are showing no signs of letting up.

In fact, Oracle spent $6.9 billion on data center infrastructure in fiscal 2024 and plans to double that figure in fiscal 2025. But it gets better.

Ellison’s latest comments are great news for Nvidia

During the meeting with analysts, Ellison told the audience about a dinner he arranged with adze CEO (and founder of xAI) Elon Musk and Nvidia CEO Jensen Huang at Nobu in Palo Alto. He remembered himself and Musk begging Huang for multiple GPUs:

Please take our money… take more of it. You don’t take enough. … We need you to take more money from us. Please.

— Ellison and Musk’s comments to Jensen Huang at the dinner, according to Ellison.

Oracle Cloud Infrastructure (OCI) generated revenue of $2.2 billion in the first quarter (primarily from leasing data center capacity to customers), which was a 46% increase over the year-ago period . However, Oracle ended the quarter with a record $99 billion in remaining performance obligations (RPO), a staggering 53% increase. The company said it signed 42 new deals for GPU capacity worth $3 billion in the first quarter, contributing to the backlog.

Oracle can’t serve all those AI developers — or turn its RPO into revenue — until it brings more data centers online, so Ellison is begging Huang for more GPUs.

Tesla is in a similar position. It’s vying for supremacy in the self-driving software industry and is trying to bring a pool of 50,000 GPUs online by the end of this year to train its AI models. Tesla will spend $10 billion on this infrastructure, but will need more capacity over time.

Image source: Nvidia.

Now could be a great time to buy Nvidia stock

Oracle and Tesla aren’t the only companies spending big on data centers. Microsoft spent $55.7 billion on capital expenditures (capex) primarily related to AI infrastructure in fiscal 2024 (ended June 30) and plans to spend even more in fiscal 2025. Similarly, AmazonIts capital spending is on track to top $60 billion this calendar year.

Based on Nvidia’s trailing 12-month earnings per share of $2.20, its stock trades at a price-to-earnings (P/E) ratio of 52.7. It is expensive compared to the P/E ratio of 30.9 Nasdaq-100 technology index, which is home to many of Nvidia’s technology peers.

However, Nvidia’s 2026 fiscal year will begin at the end of January 2025, and Wall Street expects the company to deliver $4.02 in earnings per share for the year. This places its stock at a forward P/E ratio of just 28.8. In other words, investors who are willing to hold Nvidia stock for at least the next year and a half could get a bargain at the current price — assuming Wall Street’s forecast turns out to be correct.

A slowdown in Nvidia’s activity will come eventually, as the scale of current AI spending will be very difficult to sustain over the long term. Additionally, competition is slowly setting in in the GPU space, which could erode some of the company’s market share over the next few years.

However, based on the facts at hand today, Nvidia stock is probably a good buy at the current price. AI spending from some of its biggest customers suggests a slowdown is not on the immediate horizon.