Share prices of Nvidia (NASDAQ: NVDA) are down 46% year to date. That puts Nvidia’s market cap (share price times total shares outstanding) at $390 billion at the time of writing, so the stock would need to more than double to reach a $1 trillion market cap.
At a recent investor conference, Nvidia CEO Jensen Huang provided some insights about his company’s growth opportunity with two new data centers coming out. The opportunity selling advanced chip hardware and software systems to these artificial intelligence-based factories could be enough to push Nvidia’s stock price higher over the next five years.
150,000 AI factories
Nvidia’s data center business has grown fivefold over the last four years, but the best seems yet to come. At the Bank of America Global Technology Conference on June 7, Huang told analysts, “We’re in an early adoption phase of machine learning across all of these data centers,” implying that Nvidia can grow the data center segment for a long time.
Fiscal First Quarter
OEM and other
Machine learning is a kind of artificial intelligence where a computer can use past data to predict outcomes without human input. If you’re thinking about SkyNet from the Terminator movies, you’re on the right track. It’s basically a technology that teaches a computer to write software by itself.
Thankfully, companies are not using this technology to build indestructible robots armed with weapons to think for themselves. At the Bank of America conference, Huang mentioned that natural language understanding and recommendation systems are the top applications of AI driving the biggest investment by cloud service providers.
To train a computer to understand conversational speech requires lots of data. Powerful computers running GPU accelerators, or the use of a GPU along with a central processing unit (CPU), are processing petabytes of data. The ingestion of all this data is translating to the emergence of two types of data centers.
One is an AI factory that is specifically designed to gather, process, and build data-driven predictive models. Huang sees as many as 150,000 of these AI factories being built over the long term, similar in scope to the thousands of factories that came after the industrial revolution.
Another new data center is a localized version operating at the edge. For example, self-driving cars are essentially a data center on wheels. In navigating heavy traffic, autonomous vehicles need to process an infinite number of variables in real-time to understand their surroundings.
To understand what this means for Nvidia’s long-term growth, Huang mentioned that it took 15 years for 30% of the world’s supercomputers to implement the use of GPU accelerators to process data workloads. Nvidia’s technology is currently in 80% of the new data systems coming out. This means Nvidia is positioned to be the leading hardware and systems provider for potentially thousands of new data centers.
This explains why Nvidia raised its long-term addressable market to $1 trillion at the company’s investor day in March.
What about the stock?
The stock currently trades at a price-to-earnings ratio of 24 based on next year’s analyst estimates. I would call that cheap for a company that is operating with this much momentum behind it and with a long growth runway still ahead.
Also, considering that Nvidia is moving into providing software along with chip hardware, such as Nvidia AI Enterprise and Omniverse, it should be able to grow profits faster than revenue. These platforms are sold on a license or subscription basis and provide a high-margin stream of recurring revenue. Nvidia expects at least half of its annual revenue to come from software over time.
For these reasons, it’s only a matter of time before Nvidia climbs back toward a $1 trillion market cap. I believe it can reach that level in the next five years given the leadership it has in enabling AI technology.
Bank of America is an advertising partner of The Ascent, a Motley Fool company. John Ballard has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Nvidia. The Motley Fool has a disclosure policy.