Nvidia (NVDA 0.48%) stock has been on a legendary run. Three years ago, approximately zero people could have predicted that Nvidia would become the largest company in the world. Apple (AAPL +3.26%) held that title, valued at $2.28 trillion, while Nvidia was valued at a more modest $298 billion. But after unprecedented artificial intelligence (AI) demand, Nvidia has rocketed higher and holds the title of the world's largest company by a healthy margin.
Investors can't go back in time and invest in Nvidia, but they can project out three years to see where Nvidia can go from here. I think there are two major possibilities, with one outcome being more likely than the other. Let's take a look at where Nvidia is headed, and that may clue investors in on what they should do with the stock now.
Massive AI demand will continue to drive Nvidia's stock gains
Nvidia makes graphics processing units (GPUs), which have the unique ability to process multiple calculations in parallel. This attribute makes them perfectly suited for processing arduous workloads like gaming graphics, engineering simulations, mining cryptocurrency, and training artificial intelligence models. The current generation of generative AI models has been trained on Nvidia GPUs, making them vital pieces of hardware in the AI value chain.
This has led to unprecedented revenue and profit growth for Nvidia.
NVDA Revenue (TTM) data by YCharts
With growth levels like that, investors may be wondering if there's still juice left in the tank. Luckily, there is.
AI hyperscalers have announced multiple deals with Nvidia and its partners that will once again set a record in data center capital expenditures during 2026. This bodes well for Nvidia, but investors are growing increasingly concerned that Nvidia may be losing some market share. The two biggest culprits are AMD (AMD +1.64%) and Broadcom (AVGO +0.88%). AMD recently signed a major deal with OpenAI, the maker of ChatGPT, signaling to other AI companies that AMD's products are competitive with Nvidia's, something that hasn't been the case. This could cause some companies to deploy some cheaper AMD hardware, which could result in market loss for Nvidia.
Broadcom's computing units aren't a direct replacement for Nvidia's GPUs. Instead, they make custom AI accelerator chips that are designed in collaboration with the end user. This makes them impossible to obtain if you're a relatively small company (unless you rent them through the cloud). Still, the AI hyperscalers have plenty of resources to partner with Broadcom to make these chips viable. Broadcom's custom AI chips can outperform Nvidia's GPUs at a lower cost point, but that comes at the price of flexibility, because Broadcom's chips are designed with only one workload in mind.
Both of these are threats to Nvidia's market dominance, which has been estimated to be about a 90% market share. Despite both of these factors, Nvidia is still the most popular computing unit to train and run AI models on, and it will still be in a great position for the next wave of AI spending.
AI spending is projected to reach multitrillion-dollar levels
Many investors were skeptical when Nvidia stated during its Q2 earnings call that global data center capital expenditures would reach $3 trillion to $4 trillion by 2030. However, that bold projection was quickly followed up by multiple Wall Street firms guiding for similar or increased amounts of AI spending over the next few years. This will bode well for Nvidia, especially with Nvidia estimating that global data center capital expenditures will total around $600 billion this year.
NASDAQ: NVDA
Key Data Points
That indicates a compound annual growth rate (CAGR) of about 42%, which could directly translate into Nvidia sustaining its rapid growth rate over this time frame. Should Nvidia deliver 40% growth over the next five years and maintain its current valuation, that would transform Nvidia from a $4.58 trillion business into nearly a $25 trillion giant.
A company of that size is hard to fathom, but if the AI market grows at projected rates and Nvidia maintains its market share, that's where it's heading. If that's the case, then Nvidia is a screaming buy right now. There's also the possibility that these projections are wrong and AI spending decreases, or that Nvidia loses substantial market share. I don't see the latter happening, but the former is always possible. If that's the case, Nvidia's shares will likely plummet in response to decreased AI spending.
We've barely scratched the surface of what an AI-first society looks like, making me inclined to believe that the spending projections are real. Time will tell, but if those projections pan out, then Nvidia is a must-buy at these levels.
