The pressure was on earlier this week for the most valuable company in the world to deliver an epic quarter, and it didn't disappoint. Nvidia (NVDA +0.25%) beat analyst revenue and earnings estimates, with shares surging after hours.
Here are three reasons why Nvidia remains the foundational artificial intelligence (AI) growth stock to buy now.
Image source: Getty Images.
1. Nvidia is showing no signs of slowing down
Nvidia continues to grow data center revenue and earnings at a breakneck pace, led by Blackwell graphics processing units (GPUs). Blackwell Ultra is specially engineered for large-scale data centers to boost training and inference for AI models -- which has been a key driver of Nvidia's growth. In fact, Nvidia noted that hyperscaler demand continues to outpace its supply -- signaling AI spending remains stronger than ever.
Blackwell Ultra was announced this past March at Nvidia's GTC 2025 conference. And already, Blackwell Ultra-based systems have become the leading architecture across Nvidia customer categories.
Nvidia doesn't get enough credit for its associated hardware and software that support its GPUs. The company's networking revenue grew from $3.1 billion in the third quarter of fiscal 2025 to $8.2 billion in the third quarter of fiscal 2026 -- making Nvidia the largest networking business in the world.
In sum, the insatiable appetite for Nvidia's GPUs, paired with its massive expansion on the networking side, shows that the AI investment cycle remains in expansion mode. It's worth mentioning that management is penciling in zero data center computer revenue from China for the upcoming fourth quarter. So Nvidia is delivering incredible results even while a key market is constrained.
NASDAQ: NVDA
Key Data Points
2. Margins remain high despite increasing competition
Competition is heating up as Advanced Micro Devices and Broadcom land massive cloud deals.
Broadcom works with hyperscalers to design chips for AI applications. Paired with its networking devices, like Jericho routers and Tomahawk switches, Broadcom's system can be more cost-efficient compared to general-purpose GPUs -- especially when performing tasks based on existing AI models with fixed patterns. Meanwhile, Nvidia's GPUs are the workhorses of the data center that can adapt to complex scenarios.
Despite the competition and Nvidia's increased spending, its margins are still sky-high -- proving that the pie is big enough for multiple solutions to thrive in hyperscale AI data centers.
In its latest quarter, Nvidia generated a 63.2% operating margin -- which was actually slightly higher than 62.3% a year ago.
3. Nvidia is returning cash directly to shareholders
Nvidia spent $36.3 billion on stock buybacks in the first nine months of fiscal 2026 compared to $25.9 billion during the same period in fiscal 2025.
Nvidia's stock buyback program more than offsets its stock-based compensation, thereby preventing dilution of existing shareholders. In turn, it reduces Nvidia's outstanding share count, which accelerates earnings per share growth and makes the stock a better value.
Nvidia remains a balanced AI stock to buy now
Nvidia continues to prove the AI doubters wrong by delivering quarter after quarter of breakneck earnings growth, maintaining its high margins, and generating substantial excess cash flow despite increased spending.
If Nvidia continues to deliver blowout earnings growth, it's hard to argue that the stock is expensive. However, investors should still take a long-term view when buying the stock, as an earnings miss would likely trigger a knee-jerk reaction to the downside.
