Discrete GPUs were initially designed for use in video games and professional graphics applications, but they’ve now become the backbone of data centers that use them to process all sorts of computations, especially those involving complex artificial intelligence (AI) functions. Unlike CPUs, which process a single piece of data at a time through scalar processing, GPUs use vector processing to manage a wide range of floating-point numbers and integers simultaneously.
That key difference makes powerful data center GPUs the picks and shovels of the AI gold rush. Nvidia (NVDA -0.12%) and Advanced Micro Devices (AMD 0.36%) share a near-duopoly in this market, but which chipmaker is the better AI-driven investment right now?
Image source: Getty Images.
The key differences between Nvidia and AMD
Nvidia is the 800-pound gorilla of the GPU market, while AMD is the distant underdog. Nvidia’s share of the entire discrete GPU market grew eight percentage points year over year to 90% in the third quarter of 2024, according to JPR, while AMD’s share shrank from 17% to 10%. TechInsights estimates Nvidia accounted for 98% of all data center GPU shipments in 2023.
Nvidia generated 88% of its revenue from the data center market in its latest quarter. Only 9% came from its gaming GPU segment (which was once its biggest business), while the rest mainly came from other chips for the auto and OEM markets.
AMD produces GPUs, x86 CPUs, and APUs (accelerated processing units) which merge CPUs and GPUs for gaming consoles and other devices. It controlled 38% of the x86 CPU market in the first quarter of 2025, according to PassMark Software, while Intel held a 60% share. AMD sells both types of chips for the PC and data center markets.
For PCs, AMD sells Ryzen CPUs and Radeon GPUs. For servers, it sells Eypc CPUs and Instinct GPUs. Its Epyc CPUs compete against Intel’s Xeon CPUs, while the Instinct GPUs target Nvidia’s Hopper series GPUs at a much lower price point.
However, Nvidia locks in a lot of AI developers into its proprietary CUDA (Compute Unified Device Architecture) platform for GPU applications. AMD’s GPUs can only run CUDA with non-native porting tools and applications, which further limits its appeal among large cloud and AI customers. That’s why the world’s top AI companies — including OpenAI, Microsoft, Amazon, Meta Platforms, and Alphabet‘s Google — still mainly use Nvidia’s GPUs to power their generative AI applications.
Which company is growing faster?
In fiscal 2023 (which ended in January 2023), Nvidia’s revenue growth flatlined as the PC market lapped its pandemic-driven growth spurt and the macroeconomic headwinds battered the data center market. But in fiscal 2024, its revenue soared 126% as the sudden rise of generative AI applications lit a raging fire under its data center business.
From fiscal 2024 to fiscal 2027, analysts expect Nvidia’s revenue and EPS to grow at a compound annual growth rate (CAGR) of 57% and 65%, respectively, even as it faces tighter export curbs in China, potential competition from AMD and other AI chipmakers, and possible antitrust regulations targeting its near-monopolization of the market.
It expects its near-term growth to be driven by the ongoing expansion of the AI market and the rollout of its next-gen Blackwell GPUs, which are roughly 2.5 times faster than its Hopper GPUs.
In 2023, AMD’s revenue fell 4% as the PC and gaming console markets cooled off. In 2024, its revenue rose 14% as it lapped that slowdown and sold more Epyc CPUs and Instinct GPUs for the data center market. For the full year, its data center revenue soared 94% and accounted for 49% of its top line. However, AMD’s data center growth notably slowed down in the fourth quarter and sowed doubts regarding its ability to keep pace with Nvidia in the AI-oriented GPU market.
From 2024 to 2027, analysts expect AMD’s revenue and EPS to grow at a CAGR of 20% and 73%, respectively, assuming the PC market recovers, Intel continues to struggle in the x86 CPU market, and it sells more data center CPUs and GPUs.
Which stock is the better value?
Nvidia trades at 33 times its projected earnings for fiscal 2026 (which ends in January 2026), while AMD trades at 45 times its estimated earnings for 2025. Therefore, Nvidia’s stock still looks cheaper than AMD’s relative to its growth potential — even though it has already outperformed AMD by a wide margin over the past five years.
Nvidia’s business is less diversified than AMD’s, but it generates most of its revenue from the higher-growth data center GPU market, which will probably keep expanding for the foreseeable future. AMD’s data center business is growing rapidly, but it’s still being weighed down by the cyclical PC and gaming console markets. So, for now, Nvidia is still a better value than AMD. Nvidia’s stock might run out of steam if the AI market cools off, but it should stay ahead of its smaller and messier competitor.
John Mackey, former CEO of Whole Foods Market, an Amazon subsidiary, is a member of The Motley Fool’s board of directors. Randi Zuckerberg, a former director of market development and spokeswoman for Facebook and sister to Meta Platforms CEO Mark Zuckerberg, is a member of The Motley Fool’s board of directors. Suzanne Frey, an executive at Alphabet, is a member of The Motley Fool’s board of directors. Leo Sun has positions in Amazon and Meta Platforms. The Motley Fool has positions in and recommends Advanced Micro Devices, Alphabet, Amazon, Intel, Meta Platforms, Microsoft, and Nvidia. The Motley Fool recommends the following options: long January 2026 $395 calls on Microsoft, short February 2025 $27 calls on Intel, and short January 2026 $405 calls on Microsoft. The Motley Fool has a disclosure policy.
