Nvidia (NVDA -1.44%) shares recovered nicely after the April market swoon. The stock is now just 10% off its all-time high, as of this writing. After a correction that bottomed on April 5, Nvidia shares have rocketed 48% higher. That compares to just a 27% rebound by the Nasdaq Composite since its early-April low.
Some may believe that will mean Nvidia shares could level off while the overall index catches up as other high-growth tech companies also benefit from the surge in spending for artificial intelligence (AI) infrastructure. I'd argue that Nvidia will continue to outpace its peers, though, as the next phases of AI build-out and development will also lead to surging sales and profits for Nvidia.
Image source: Getty Images.
Nvidia's next big growth driver
The first stage of the AI wave has been adding massive amounts of computing power to build and train models. Nvidia's sales exploded fourfold in its data center segment in just the last two years, materially by selling advanced graphics processing units (GPUs) needed for those powerful computer clusters.
As the large language models (LLMs) built with that chip power multiply, another phase in the AI revolution is taking center stage. Nvidia CEO Jensen Huang called it out in the company's recent fiscal 2026 first-quarter report, released May 28. Huang said:
Global demand for Nvidia's AI infrastructure is incredibly strong. AI inference token generation has surged tenfold in just one year, and as AI agents become mainstream, the demand for AI computing will accelerate.
AI inference refers to the practical application of using LLMs in the real world. The abundance of LLMs now include OpenAI's ChatGPT and Anthropic's Claude, as well as many AI-powered search engines and chatbot platforms using multiple LLMs. Private and public enterprises are also running their own custom models tapping into LLMs.
The inference tokens Huang refers to are units of data -- groups of text that the models read, process, and generate. That occurs after any question is answered, or an AI agent does a task. Tokens need a query or prompt when they are processed by GPUs.
The volume of those queries is unsurprisingly growing and becoming more complex. The more complex the interaction, the more tokens it consumes. As with AI training, Nvidia is dominant in inference.
AI factories are the next era of artificial intelligence
Nvidia is now leading the global development of AI factories. That's a smart strategy that helps take advantage of the entire AI ecosystem by keeping it in-house as much as practical. AI factories are effectively highly advanced data centers.
AI factories do more than just store and process data. They host the full life cycle of AI from data input, training, tweaking, and calibrating, to the voluminous amount of inference. The result gives enterprises the ability to enable prediction, generate content, and perform reasoning to seek solutions.
Nvidia's products play roles in all of these functions. The company supplies software that optimizes its most advanced GPUs to lower per-token costs for customers. Nvidia offers an example where integrating software optimizations and adopting the latest-generation chips reduced costs by up to 20-fold versus older GPUs and processes. Customers can thus see more material returns on investment.
The market is global for Nvidia. It is partnering with Saudi Arabia's state-owned AI company Humain to build AI factories that will utilize hundreds of thousands of Nvidia's most advanced GPUs as they are developed in the next several years.
Nvidia is an investment for the future
Nvidia realized a massive amount of success over the last several years as AI technology entered the mainstream with chatbots. Plans to build infrastructure blossomed, and Nvidia was the primary supplier able to support what was necessary to build and train models.
More services are now being built around the LLMs, too. Nvidia has software stacks to support the models and supercomputers along with related services. Competition may rise around GPU supply, but Nvidia has many more products to continue growing revenue as AI expands across virtually every sector.
Investors owning Nvidia stock now should have a solid chance of beating the market for years to come.
