Nvidia (NVDA +1.54%) is one of the market's most talked-about stocks. The chipmaker's shares soared 1,220% over the past five years as it evolved from a maker of gaming graphics processing units (GPUs) into the gatekeeper of the booming artificial intelligence (AI) market.
Nvidia achieved that evolution by developing high-end GPUs for accelerating AI tasks at data centers. Unlike central processing units (CPUs), which process one piece of data at a time, GPUs can process a broad range of data points simultaneously. That's why the top generative AI platforms -- including OpenAI's ChatGPT and DALL-E -- are currently powered by Nvidia's GPUs.
Nvidia CEO Jensen Huang. Image source: Nvidia.
The AI market's explosive expansion over the past few years sparked a buying frenzy in Nvidia's data center GPUs. That growth offset the lumpier growth of its gaming GPU business, which accelerated throughout the pandemic but faced a tough slowdown as the PC market cooled off. In its latest quarter, Nvidia generated 80% of its revenue from its data center chips and just 16% of its revenue from its gaming business.
Analysts expect Nvidia's revenue to soar 118% in fiscal 2024 (which ends in January 2024) and grow at a compound annual growth rate (CAGR) of 35% from fiscal 2024 to fiscal 2026. Those growth rates are stunning, but is it a good long-term play?
The bull case is straightforward
As the clear leader in high-end GPUs for data centers, Nvidia is poised to profit from the long-term expansion of the AI market. Precedence Research expects the broader AI market to grow at a CAGR of 19.1% from 2023 to 2032. Within that market, Fortune Business Insights believes the higher-growth generative AI niche will expand at a CAGR of 47.5% from 2022 to 2030.
During Nvidia's latest conference call in November, CFO Colette Kress said investments in "infrastructure for training and inferencing large language models, deep learning recommender systems, and generative AI applications is fueling strong broad-based demand" for its data center GPUs. Kress also said Nvidia had "significantly increased" its supply "every quarter this year" and will continue to do so next year" to "meet a growing and diverse set of AI opportunities."
Nvidia also remains far ahead of Advanced Micro Devices in the gaming GPU market. It controlled 87% of the discrete GPU market in the second quarter of 2023, according to JPR, compared to AMD's meager 10% share. Therefore, the development of more graphically intensive games should continue to drive the growth of its gaming business.
All these catalysts suggest Nvidia's stock still has plenty of room to run -- and it doesn't look too expensive at 24 times forward earnings and 13 times next year's sales.
NASDAQ: NVDA
Key Data Points
The bear case is more complicated
Nvidia's strengths are easy to see, but it also has three glaring weaknesses. First, the AI market might be due for a breather from its feverish growth over the past few years. That slowdown could result in a supply glut and lower GPU prices.
Second, there's no guarantee Nvidia will maintain its early-mover advantage in the data center GPU market over the next decade. OpenAI, its biggest backer Microsoft, Meta Platforms, and Alphabet's Google have all been developing their own first-party AI accelerators to curb their dependence on Nvidia's chips. Smaller companies like Graphcore have also been trying to develop more cost-efficient AI accelerator chips.
Lastly, the escalating tech war between the U.S. and China could curb Nvidia's long-term growth. The U.S. government already blocked Nvidia from selling its top-tier A100 and H100 data center chips to Chinese customers last year, but Nvidia shrewdly offset that by selling more of its lower-end A800 and H800 chips in China instead. However, the U.S. closed that loophole in October, banning those sales as well.
Nvidia believes it can mitigate that by diverting its China-bound chips to other markets, but that's only a short-term solution to a longer-term problem. The tighter export restrictions could drive Chinese chipmakers to develop their own homegrown AI chips to replace Nvidia's technology -- and those chips could eventually seep into other overseas markets.
Is Nvidia a good long-term investment?
Nvidia isn't a bulletproof investment, but I believe its strengths outweigh its weaknesses. The AI market could cool off, but it still has plenty of room to grow as companies adopt more AI tools to optimize and accelerate their operations.
Nvidia should also generate plenty of cash to develop new chips before any meaningful competitors actually arrive in the data center GPU market, and its scale should prevent those rivals from gaining much ground as the AI market expands. Its future in China remains unclear, but it can continue to grow in other markets that aren't as exposed to the trade tensions between the U.S. and China. It will also likely take Chinese chipmakers a long time to develop comparable AI chips on their own.
Therefore, Nvidia's rally might stall out as the AI fever dies down, but it's still a sound investment for long-term investors who can tune out all the near-term noise.
