AMD Surpasses Intel in Data Center Chip Revenue

Intel's Xeon processors have long been the backbone of most servers, with AMD's processors holding only a small single-digit market share around seven to eight years ago. However, a significant shift has taken place in recent years. While Intel's Xeon CPUs continue to dominate the majority of server installations, high-end devices are increasingly turning to AMD's EPYC processors, which has contributed to AMD's data center business surpassing Intel's Data Center and AI Group (DCAI) in revenue.

In fact, AMD's data center segment generated $3.549 billion in revenue during Q3 2024, compared to Intel's DCAI division, which earned $3.3 billion during the same period. Just two years ago, Intel's data center and AI business was generating between $5 billion and $6 billion in revenue per quarter. This drop in Intel's earnings can be attributed to the growing competitive edge of AMD's EPYC processors over Intel's Xeon CPUs, prompting Intel to sell its server chips at steep discounts, which has impacted both its revenue and profit margins.

Intel's flagship 128-core Xeon 6980P "Granite Rapids" processor, priced at $17,800, is the company's most expensive standard CPU to date. In contrast, AMD's highest-end 96-core EPYC 6979P processor costs $11,805. While demand for Intel's Xeon 6900 series remains high, and the company could potentially stabilize its data center revenue if it increases production of the Granite Rapids chips, it still faces challenges in scaling up supply.

Although Intel and AMD each generate approximately $3 billion to $3.5 billion per quarter from sales of data center CPUs, NVIDIA's profits from its data center GPUs and network chips far exceed those of both companies. These chips are essential for enabling AI processors to work together in data center environments. In fact, for Q2 FY2025, NVIDIA's network product sales reached $3.668 billion, while its compute GPU sales topped $22.604 billion—well above the combined sales of Intel and AMD's data center hardware. Overall, NVIDIA's AI and HPC (high-performance computing) GPU sales for the first half of 2024 reached nearly $42 billion, with the company expected to continue seeing strong sales in the second half of the year.