All this is of great interest to gamers like my son, who asked all weekend whether he should plunk for a new graphics engine on his 3-year-old PC. It may even stimulate near-term volatility and profits for short sellers.
But if you’re a long-term investor and see strength 10 years ahead of you instead of infirmity, this is not where the game is. The game for you is artificial intelligence. It’s in the clouds and new devices we can barely imagine today.
That’s the game Nvidia aims to win.
The AI Game
Nvidia has been focused on the cloud market in recent years, specifically in general purpose chips that accelerate the work of clouds. They dominates that market the way Intel (NASDAQ:INTC) once dominated PCs.
Rivals are hiding in niches, trying to turn what had been field-programmable gate arrays (FGPAs) into competition for general purpose Nvidia chips. Xilinx (NASDAQ:XLNX), Intel and AMD are all doing similar things. But Nvidia still has the most complete solution for cloud-based artificial intelligence (AI). That will become more apparent once its acquisition of Mellanox (NASDAQ:MLNX), whose “networking fabric” delivers optical speeds within data centers, is complete.
Nvidia’s recent construction of its own AI computer, dubbed the SuperPod, demonstrated just how big its data center ambitions are.
Nvidia’s competition won’t come from other chipmakers but from the cloud giants themselves. Alphabet (NASDAQ:GOOGL, NASDAQ:GOOG), Microsoft (NASDAQ:MSFT) and Amazon (NASDAQ:AMZN) are all working on AI chip sets for their clouds, knowing that keeping costs down on scaled infrastructure is still their key to success.
To win them over Nvidia is demonstrating inference chips with 32 “chiplet” processors, while commercial chips are running with four or eight processors. Nvidia’s processor is like a cloud on a chip.
None of this is cheap. Nvidia’s research budget last year came to nearly $2.4 billion, one-third more than the previous year. That’s twice as much as it spends on selling and general administration, over one-third of total operating spending. But if it keeps Nvidia ahead of its customers, it’s money well-spent.
NVDA Has Much to Win
There’s a common misperception among investors that everything falls in a recession.
It’s not true. Throughout this century technology has recovered ahead of the general economy. Technology that can cut costs, not just create new markets, has done especially well.
That’s where Nvidia expects to win. Replacing labor, even intellectual labor, with machines boosts productivity once customers realize they must either innovate or die. It eliminated middle managers in the 2000s, insurance salesmen in the 2010s, and is bound to eliminate millions of drivers and machine operators in the next decade.
What will be left will be the value those jobs provided. Deliveries will still reach you, just as you can still buy insurance and run a scaled enterprise. A recession that other companies anticipate with dread, Nvidia approaches with hope.
The Bottom Line for Nvidia Stock
Nvidia’s 0.39% dividend yield is not the reason you buy the stock. Income investors should look elsewhere.
Nvidia’s short-term growth prospects are also poor. It brought in $2.22 billion during the first quarter of its 2020 fiscal year. It’s expected to bring in just $2.55 billion this quarter, which will be reported August 15. At its current pace, it will be tough to match last year’s $11.7 billion.
You buy Nvidia with your eyes on the horizon — three years, five years, 10 years out. CEO Jensen Huang is 56. He should be around for the next decade.
He’s a name worth betting on.
Dana Blankenhorn is a financial and technology journalist. He is the author of a new environmental story, Bridget O’Flynn and the Bear, available now at the Amazon Kindle store. Write him at firstname.lastname@example.org or follow him on Twitter at @danablankenhorn. As of this writing he owned shares in AMZN and MSFT.