Quick Buzz Feed
The growth of AI inference workloads in data centers is boosting demand for server CPUs, a market that's dominated by AMD and Intel.
Artificial intelligence (AI) computing power, traditionally focused on training large language models (LLMs), is now shifting significantly towards inference. Inference involves applying trained AI models to real-world data and queries. By 2030, McKinsey predicts that inference will constitute over half of AI computing capacity in data centers, surpassing the training phase. This shift is driving a robust increase in demand for central processing units (CPUs), as they are considered cost-effective for handling AI inference workloads. Advanced Micro Devices (AMD) recently doubled its server CPU total addressable market (TAM) estimate to $120 billion by 2030, attributing this growth to agentic AI and inference applications.
Intel currently holds the dominant position in the server CPU market, with approximately 67% market share at the end of Q1 2026, according to Mercury Research. AMD accounts for the remaining market share. However, Intel has been consistently losing ground to AMD over the past four years, having controlled over 88% of the market in Q1 2022. AMD's success is attributed to offering superior products that attract more customers, and it enjoys stronger pricing power, as evidenced by its record 46.2% revenue share in Q1, indicating higher average selling prices for its server processors. AMD anticipates significant growth, projecting a 70% increase in server CPU revenue in Q2 2026, with robust growth expected to continue through 2026 and into 2027 with its next-generation EPYC processors. Intel, on the other hand, faces challenges with short supply of its Xeon server CPUs, although demand remains strong. Intel aims to counter AMD's gains by ramping up production of its 18A architecture-based server CPUs, which promise substantial performance and efficiency improvements over AMD's Epyc processors. Intel is also partnering with Nvidia to integrate Xeon 6 processors into its Vera Rubin server racks, indicating a strategic move to maintain its market position.
Both Intel and AMD stocks have seen substantial appreciation this year, with Intel up 216% and AMD up 129% at the time of writing. Despite these gains, AMD appears to be a more attractively valued option. Consensus estimates suggest that AMD's earnings growth will significantly outpace Intel's, with AMD's bottom line expected to grow by nearly 78% in 2027, compared to Intel's projected 42% growth. Furthermore, AMD's increasing influence in the data center GPU market, securing lucrative contracts with major hyperscalers and AI companies, provides it with a distinct advantage in the broader AI chip sector. Intel has struggled to gain significant traction in the data center GPU market, reinforcing AMD's stronger position in AI. Therefore, AMD's relatively cheaper valuation, superior earnings growth projections, and strong presence in the GPU market make it the preferred AI stock to capitalize on the expanding inference demand.
While Advanced Micro Devices is a strong contender in the AI inference market, investors are encouraged to consider diverse investment strategies. The Motley Fool's 'Stock Advisor' analyst team, known for consistently outperforming the S&P 500, did not include Advanced Micro Devices in their list of top 10 stocks to buy now. This suggests there are other opportunities for long-term growth that could yield significant returns. For example, previous recommendations for Netflix and Nvidia in 2004 and 2005, respectively, demonstrated substantial returns. Investors seeking to capitalize on such opportunities are advised to explore 'Stock Advisor's' latest recommendations for a diversified, long-term growth portfolio.