Perhaps the most important piece of technology powering generative AI development is an advanced chipset known as the graphics processing unit (GPU).
While Nvidia pioneered these chips, two of the company's cohorts in the semiconductor space -- Advanced Micro Devices (AMD 1.11%) and Intel (INTC +2.53%) -- are seeking to make inroads as investments in artificial intelligence (AI) infrastructure accelerate.
Let's break down the current picture between AMD and Intel, and assess which chipmaker is better positioned for the proliferation of the AI infrastructure era.
Image source: Getty Images.
Where does AMD's data center business stand today?
The AI revolution has featured a number of different GPU series. For Nvidia, the company's Hopper architecture was once coveted as the best chips money could buy. But over the last couple of years, Nvidia has introduced successor chips -- namely, Blackwell and the upcoming Rubin architecture.
Similarly, AMD started to gain traction in the data center landscape through its Instinct MI300 accelerators, launched in the fourth quarter of 2023.
As the graphic below illustrates, AMD's data center business generated similar levels of revenue to Intel's within about six months following the Instinct release.
Image source: The Motley Fool.
During the third quarter of 2025, AMD's data center segment generated $4.3 billion in revenue -- rising by 22% year over year. By contrast, Intel's data center business reported $4.1 billion of sales -- a decline of 1% annually.
NASDAQ: AMD
Key Data Points
Could Intel be a stealth data center stock?
Intel is an extremely diversified business. In addition to its data center operations, the company also sells various hardware products, including microprocessors and other chips, as well as provides foundry services.
A couple of months ago, Nvidia agreed to invest $5 billion into Intel alongside additional funding commitments from the U.S. government and SoftBank. As part of the deal, Intel will be designing next-generation CPU architectures for Nvidia -- a potential catalyst for its data center segment.
The verdict: AMD's full-stack approach beats Intel's diversification
As the image above illustrates, AMD's data center segment has consistently grown at double-digit rates. By contrast, Intel's growth has been rather inconsistent and appears to be decelerating in more recent quarters.
I believe the growth disparity between the two chipmakers stems from their differing approaches.
AMD offers a combination of hardware -- GPUs and CPUs -- as well as a software system called ROCm (Radeon Open Compute) -- providing developers with a comprehensive, full-stack suite. This approach is similar to that of Nvidia, which complements its GPUs with a custom software platform called CUDA.
By offering both hardware and software, AMD is able to create a lock-in effect with its developer base -- putting the company in a position to scale alongside its customers as they continue to invest capital in AI capex.
This strategy has already helped AMD win over the likes of Microsoft, Meta Platforms, Oracle, and OpenAI -- each of which is deploying large clusters of the company's Instinct accelerators.
Conversely, Intel has been struggling to execute on the innovation front. To further exacerbate its operational headaches, most of the hyperscalers now turn to Taiwan Semiconductor for its foundry services instead of Intel.
Throughout the AI revolution, TSMC has seen its market share in the foundry space expand from 56% to 68%. Meanwhile, Intel has lost ground -- now accounting for less than 1% of the market.
While Intel has some intriguing catalysts in the works, it's hard to know just how impactful its relationship with Nvidia will be. To me, Intel remains more of a turnaround story than a concrete winner of the AI infrastructure chapter.
For these reasons, I see AMD as the more strategically positioned data center business. The company's progress to date appears to be challenging an incumbent like Intel, and its current momentum on the backdrop of high-profile customer wins and secular themes of a multi-year, multi-trillion dollar infrastructure opportunity could further widen the gap.
