Nvidia (NASDAQ: NVDA) has been one of the best-performing stocks in the market since early 2023, but some investors are starting to wonder if its incredible run is running out of steam. I’m not in that group. The economy is still in the early stages of what is shaping up to be a generational shift, and Nvidia is right at the heart of it.
There are plenty of reasons investors should hang on to their Nvidia shares (or buy more). But I want to focus on three that have me convinced that Nvidia is still a must-own stock.
In terms of overall excellence, Nvidia’s graphics processing units (GPUs) are best in class, which explains why its products are so dominant. GPUs are designed to handle tough computing tasks due to their ability to process multiple calculations in parallel. This gives them a leg up in processing tasks like artificial intelligence (AI) model training. In a separate but related issue, GPUs have been integral to facilitating the move from on-premise computing to cloud computing, powering the data centers that make cloud computing work.
The buildout of AI and the data centers to facilitate AI processing is scaling up at a fast pace, but recent concerns about a global economic slowdown and rising uncertainty related to ever-changing tariff policies has left some investors wondering what effects this will have on Nvidia.
At the moment, most AI hyperscalers have confirmed plans to spend record amounts on building data centers, which is excellent news for Nvidia. Even better, Nvidia expects this increased spending to persist over multiple years.
Using third-party data, Nvidia estimated that about $400 billion was spent on data center capital expenditures in 2024. Nvidia management expects that to rise to $1 trillion by 2028. Because of their size, data centers take years of planning to build, so growth of this magnitude over the next few years is quite plausible.
Data centers have grown to become Nvidia’s biggest business segment, and it looks like it will grow even bigger going forward.
Artificial intelligence usage has certainly ramped up over the past few years, but it’s nowhere near peak usage. Businesses worldwide are still working on integrating AI into their operations, which will require more computing power to support the increased workloads.
