Nvidia (NASDAQ: NVDA) is up a mind-numbing 21,840% over the last decade. At first glance, it might seem like a stock that was a generational buy rather than one that can still deliver incredible returns. The same goes for its “Magnificent Seven” peers.
Here’s why Nvidia still has a long runway for growth and why it remains a generational buying opportunity hiding in plain sight.
|
Will AI create the world’s first trillionaire? Our team just released a report on a little-known company, called an “Indispensable Monopoly,” providing the critical technology Nvidia and Intel both need. |
In October, Nvidia became the first company to surpass $5 trillion in market capitalization, reaching an all-time intraday high and closing at a high on Oct. 29. But roughly six months later, the stock remains down by about 4% from that peak at about $4.85 trillion, even though Nvidia has continued to grow earnings, innovate, and raise its long-term guidance.
Nvidia now says it expects at least $1 trillion in artificial intelligence (AI) chip revenues in 2026 and 2027. The stagnating stock price, paired with optimistic earnings expectations, has pushed Nvidia’s forward price-to-earnings ratio down to just 24.
Nvidia is already a reasonable value, but it could end up being dirt cheap in hindsight if its AI roadmap comes to fruition.
Nvidia’s best quality is its flexibility — an especially rare attribute for a company of its size.
It has, over the years, pivoted from a professional visualization and gaming company to primarily providing AI chips for data centers. But its data center business has been anchored by general-purpose graphics processing units (GPUs) for training AI models. Hyperscale data centers have far higher energy and compute requirements, demanding energy and cost efficiency across AI chips, networking, and IT equipment. On its March earnings call, Nvidia competitor Broadcom said it believes its custom application-specific integrated circuits, which are designed for narrower workloads, will eventually overtake traditional GPU designs in data centers — a threat to Nvidia.
Nvidia has wasted no time addressing this concern. It has developed a rack-scale solution under its Vera Rubin architecture that includes a GPU, a central processing unit (CPU), memory chips, and interconnects to achieve what Nvidia calls extreme co-design. The system is purpose-built for the age of AI inferencing, which uses AI models on previously unseen data, such as autonomous driving or AI agents. Nvidia is betting big on the widespread adoption of AI agents and a boom in demand for AI inference tokens — the currency needed to pay for AI usage. Nvidia’s hardware and software are built to process tokens as fast as possible, which is appealing to its hyperscale customers.
