This article was originally published on Fool.com. All figures quoted in US dollars unless otherwise stated.
Some of the best companies in the world have gotten caught up in the recent market sell-off, including Nvidia (NASDAQ: NVDA). The stock is trading down more than 25% off its highs set earlier this year as of this writing.
However, this weakness has created a nice buying opportunity in the stock. Let's look at three reasons why investors should buy the stock like there is no tomorrow.
1. Nvidia is a cheap growth stock on sale
The recent stock market pullback has left Nvidia at a very attractive valuation. It now trades at a forward price-to-earnings ratio (P/E) of only 24.5 based on this year's analyst estimates. That's not all, though; its price/earnings-to-growth (PEG) ratio is below 0.5. PEG ratios take into consideration a company's growth, and a multiple under 1 is generally considered undervalued. As such, Nvidia's stock is extremely cheap when you factor in its growth outlook.
Data by YCharts.
Nvidia has been one of the premier growth stocks in the past couple of years. Its revenue has skyrocketed 383% in the past two years, going from $27 billion in fiscal 2023 to $130.5 billion in fiscal 2025 (ended January). It's not often you see a company the size of Nvidia double its revenue in a year and then double it again the next year.
Meanwhile, its strong revenue growth is expected to continue in 2025. Analysts expect the company to grow its revenue this year by another 54% to $204.4 billion and then another nearly 24% in 2026 to $252.4 billion. While Nvidia hasn't given any full-year guidance, it did forecast its fiscal Q1 revenue to jump by 65%.
2. Nvidia's business has a wide moat
Nvidia is the clear leader in graphic processing units (GPUs), with an over 80% market share. GPUs were originally designed to speed up graphics rendering in video games. However, today, they have become the backbone of artificial intelligence (AI) infrastructure due to their ability to process vast amounts of data quickly. Along with their high memory bandwidth, this makes the chips perfect for handling the rigors of AI workloads.
Nvidia, meanwhile, has set itself apart in the field with its CUDA software platform, which it created to allow its chips to be programmed for tasks outside their original purpose. Its closest competitor, Advanced Micro Devices, didn't come out with its own software solution to program its chips until about a decade later, giving Nvidia a big lead in this area. Meanwhile, the company has since built a collection of libraries and tools on top of CUDA, called CUDA X, that now allows developers to better utilize its software for AI applications.
The moat that Nvidia has created was on display in recent testing done by semiconductor research company SemiAnalysis, which found AMD's GPUs unusable out-of-the-box for AI training due to bugs in its software and that this took away any on-paper advantages its GPUs may have had. Meanwhile, it noted that Nvidia's GPU's out-of-box performance was "amazing" with little need for any technical support from Nvidia engineers.
3. Nvidia is the AI infrastructure winner
Given its wide moat, what really matters most for Nvidia at this point is AI infrastructure spending. The company has a dominant market share, so as long as AI data center spending continues to rise, the company is well-positioned to continue to grow.
Right now, the only way to significantly improve AI models is through training them on larger and more diverse datasets. This requires more computing power, which comes from AI chips, particularly GPUs. As AI models have advanced, there has been a steep upward curve in the number of GPUs used to train them. Meta Platforms' Llama and xAI's Grok models are two of the best publicly available examples, with the latest versions of both models using 10 times as many GPUs to be trained on than their predecessors.
Meanwhile, cloud computing companies are among the biggest buyers of GPUs as they look to expand capacity to keep up with AI workload demand. In addition to building their own AI models, these companies essentially provide AI as a service where they give access to pre-trained foundation models, machine learning frameworks, and APIs, which developers use to integrate AI into their existing apps. The three big cloud computing companies will spend around $250 billion expanding their data center infrastructure this year in aggregate to keep up with the growing demand for these services.
In addition, other tech companies, such as the aforementioned Meta and xAI, are also spending big on AI infrastructure developing AI models, while OpenAI, the maker of ChatGPT, has announced plans along with Softbank and others to spend $500 billion on data center capital expenditure over the next few years through the Stargate Project. Large enterprises in various industries, meanwhile, are also investing in AI infrastructure, often taking on a hybrid cloud strategy where they use both on-premise infrastructure and cloud services.
For its part, Nvidia sees AI data center infrastructure spending topping $1 trillion in 2028. If that happens, Nvidia's stock has plenty of upside from here.
This article was originally published on Fool.com. All figures quoted in US dollars unless otherwise stated.
