Jennifer Saibil, The Motley Fool
5 min read
In This Article:
-
Nvidia has an extremely high percentage of the market for artificial intelligence (AI) chips.
-
It continues to report very high growth and strong profits.
-
Nvida is trading at an attractive price relative to next year's earnings.
Nvidia (NASDAQ: NVDA) has delivered extreme wealth for early investors, but it's also rewarded investors who bought shares fairly recently. Although it's been under pressure this year, it's up more than 740% over the past three years. That's far more than double.
Buying Nvidia stock today would be a no-brainer for anyone who thinks it can replicate that success. But it wouldn't even have to achieve such a rare feat to be valuable to investors. If it doubles five years from now, it would likely beat the market and be a very valuable addition to almost any portfolio.
Let's see if that might be in the cards.
Nvidia has had an unbelievable run over the past few years since generative artificial intelligence (AI) became the hottest thing since sliced bread. Tech companies are racing to develop the most competitive AI platform or integrate some aspect of AI into their user interfaces. Companies now offer AI-generated content, images, summaries, and more, and they're using agentic AI to get more stuff done with less human intervention.
What's common to all of these applications is that they need powerful chips to make the magic happen.
There are two basic parts of bringing generative AI to life. One is training, or the input stage, where large language models (LLM) scan tons of data to build their knowledge base. The other is inference, where the LLMs "infer" what's being asked or prompted to deliver the required output. Nvidia also services other industries with its powerful platform, including gaming, which is what drove higher sales before generative AI heated up.
Nvidia has been unstoppable. In the fiscal 2026 first quarter (ended April 27), sales increased 69% over last year. Earnings per share (EPS) were $0.81, including a charge it took for shipments it couldn't send to China because of U.S. policy. Even with the $0.15-per-share impact, Nvidia's EPS were well above last year's figure.
As LLMs become more stable and effective, they're now going past inference to reasoning, which is a more developed stage of inference that takes a bit longer, but produces better results. Instead of popping out instant answers, LLMs that use reasoning might take a few seconds to "think" and then respond.
