The next big theme in the market is quickly becoming agentic artificial intelligence (AI). There are several ways to play this trend, both with semiconductor stocks and software-as-a-service (SaaS) stocks.
With the rise of AI agents, AI data centers will need servers that incorporate a lot more central processing units (CPUs). While graphics processing units (GPUs) were ideal for providing the raw compute power to train AI models, AI agents require chips that can handle sequential reasoning and work with tools. That’s where CPUs come in, which act as the brains of a computer.
With large language model (LLM) training, the typical ratio of GPUs to CPUs was 8:1, while with agentic AI, that ratio moves to 1:1. At the same time, agentic AI works best with high-performance CPUs with high core counts that act like individual workstations. High core counts should also lead to higher CPU prices. The data center CPU market is expected to explode in the coming years, with Nvidia recently predicting this would become a $200 billion market in the next several years.
Meanwhile, on the software side, AI agents need to work with clean, structured data. Otherwise, there is a risk that they could hallucinate and make very costly errors. This opens the door for strong growth on this side of the market, as well.
Let’s look at three growth stocks to split $10,000 between to ride the agentic AI wave.
AMD: The data center CPU leader
Today’s Change
(3.93%) $20.48
Current Price
$542.02
Key Data Points
Market Cap
$885B
Day’s Range
$524.30 – $546.32
52wk Range
$114.71 – $546.44
Volume
18.7K
Avg Vol
37.5M
Gross Margin
47.09%
As a data center CPU leader, Advanced Micro Devices (AMD +3.93%) is well-positioned to benefit from surging demand due to the rise of agentic AI. The company has long been taking market share in the space from rival Intel, and it appears well ahead with its technology roadmap.
AMD’s newest Venice CPU is now beginning mass production. It uses Taiwan Semiconductor Manufacturing‘s advanced 2-nanometer process technology, supporting the use of advanced memory, such as Low-Power Double Data Rate (LPDDR), and delivering greatly improved power efficiency. The processor also packs in 256 cores, making it ideal for agentic AI.
AMD is well-positioned to be an agentic AI winner, and Venice should help the company to continue to take share in the space.
Arm Holdings: The CPU upstart
Today’s Change
(2.19%) $8.82
Current Price
$411.53
Key Data Points
Market Cap
$438B
Day’s Range
$373.87 – $417.33
52wk Range
$100.02 – $427.99
Volume
2.3K
Avg Vol
9.7M
Gross Margin
94.08%
Arm Holdings (ARM +2.19%) has long provided the intellectual property (IP) architecture to help customers design their own custom CPUs. Among its customers in this area are Nvidia, Amazon, and Alphabet, and as these companies ramp up the production of their chips, it’s set to see a nice benefit.
However, Arm surprised investors when it announced that, for the first time, it would design its own chips, targeting the data center CPU market. The company projected that it could achieve $25 billion in revenue in 2031, a sixfold increase from 2025. Approximately 15% of this would come from its new CPUs, and it believes it can take a 15% market share in what, at the time, it viewed as a $100 billion market. However, Nvidia has recently called this a $200 billion market, so its future revenue estimate could be low.
With it set to benefit from its own chips and custom CPUs, it looks like it will be a nice long-term agentic AI winner.
Image source: Getty Images.
Palantir: The orchestration leader
Today’s Change
(-6.52%) $-9.92
Current Price
$142.25
Key Data Points
Market Cap
$341B
Day’s Range
$141.25 – $151.66
52wk Range
$118.93 – $207.52
Volume
5.9K
Avg Vol
46.1M
Gross Margin
84.07%
Hardware isn’t the only way to play the rise of agentic AI, and on the software side, Palantir Technologies (PLTR 6.52%) looks uniquely positioned to benefit. The company’s AI platform (AIP) has already been seeing huge momentum, and it is perfectly situated to play an important role in agentic AI orchestration.
The success of AIP stems from the platform’s ability to gather data from disparate sources and structure it into an ontology that it links to physical assets and real-world processes. While other SaaS companies have developed platforms that can provide AI agents with the clean, structured data they need, AIP’s ontology takes it up a level, as it allows AI agents to understand how a business actually works. Palantir is already seeing extraordinary growth, and with the rise of agentic AI, that growth looks poised to continue.
As one of the biggest agentic AI winners on the software side, Palantir is a stock to own for the long haul.
