- ARM aims to achieve 50% share of CPUs in data centers by 2025.
- The rise of the Artificial Intelligence drives demand for efficient processors.
- Companies like AWS, Google, Microsoft and NVIDIA already use ARM chips.
- ARM's licensing model and energy efficiency position it as an alternative to x86.
ARM architecture is gaining ground in a sector traditionally dominated by Intel and AMD: data centers. With the arrival of new technological demands focused on artificial intelligence, energy efficiency has become a key value, and ARM wants to capitalize on this paradigm shift. The British company aims to achieve an ambitious goal: to control half of the global server CPU market by the end of 2025.
This move represents a significant leap from the 15% market share the company held at the end of 2024, according to data provided by Mohamed Awad, head of infrastructure at ARM. With AI growing at an accelerated pace and data centers multiplying their processing needs, ARM is firmly committed to a clear competitive advantage: the low power consumption of its chips. To learn more about how different servers work, you can read about servers in PUBG , which also demonstrates how optimized designs are used.
Ambition backed by the growth of AI and the quest for efficiency
The rise of artificial intelligence is driving demand for more sustainable and scalable computing solutions, directly benefiting ARM. Compared to traditional x86 architectures, such as those used by Intel and AMD, ARM-based chips stand out for their low power consumption without sacrificing performance. This is especially attractive to technology companies that manage massive infrastructures, such as large cloud service providers. If you're interested in how to change DNS settings in Windows 11 to improve your connection, you can consult this article.
The growth of servers specifically designed for AI-related tasks—such as inference and training of complex models—demands superior energy efficiency, according to Awad. Based on this premise, ARM is positioning its Neoverse platform as an optimized solution for this emerging market, which has attracted the interest of giants like Amazon Web Services (AWS), Microsoft, Google, and NVIDIA.
In addition to offering technical advantages, ARM's business model also contributes to this momentum: the company designs the chips, but doesn't manufacture them. Instead, it licenses them to third parties like TSMC or Samsung, allowing it to scale rapidly without requiring large investments in production infrastructure.
Growing adoption by major cloud players
Companies like AWS, Microsoft Azure, and Google Cloud have already begun developing or deploying ARM chips in their data centers. In Amazon's case, its Graviton processor family already represents a significant proportion of the company's server fleet. It is estimated that more than half of the chips they have added in recent years are based on this architecture.
Google has followed this trend with its Axion chip, also based on ARM Neoverse V2, while Microsoft has introduced its Cobalt 100 CPUs, also developed in-house using ARM architecture. These moves reflect a strategic shift by major tech companies toward more efficient and customized solutions, giving ARM a significant boost within the cloud ecosystem. This is similar to the importance of servers in other contexts, such as DLNA servers.
NVIDIA has also integrated Grace processors with ARM architecture into its product lineup, specifically designed to handle the complex configurations required in AI training systems. DGX servers equipped with these chips are expected to see strong demand this year, further fueling ARM's growth in this segment.
An ambitious goal but with strategic foundations
The increase from the current 15% to a 50% market share in just one year may seem aggressive, and some analysts consider this forecast more aspirational than realistic. Firms like Omdia believe that figure will be closer to a range between 20% and 23% by the end of 2025, based on global server shipment projections.
Even so, the adoption rate by so-called 'hyperscalers'—AWS, Google, Microsoft, and Alibaba—could favor ARM in this endeavor. It is estimated that AWS alone could deploy more than 1,2 million ARM CPUs in its servers during 2025, underscoring the strong momentum of this architecture in mission-critical environments.
It's not just about traditional CPUs. Many complementary devices in data centers, such as SmartNICs, DPUs (data processing units), and management cards like Nitro, also use ARM cores, further expanding the presence of this architecture in the server ecosystem.
An architecture that goes beyond the server
ARM's interests aren't limited to servers. The company plans to expand its presence into other areas such as Windows PCs , connected cars, mobile devices , and wearables. In fact, while its previous expectations were to capture 50% of the PC market by 2029, it's now focusing more on consolidating its position in data centers, where the landscape is more favorable due to energy demands and the growth of AI.
Rene Haas, CEO of ARM, stated that their ambitions in artificial intelligence extend far beyond data centers. They also aim to encompass critical parts of the network, automotive applications, smart home devices, and other elements of what is known as the “edge,” where proximity to the end user demands low latency and high efficiency. The versatility of their architecture is similar to that discussed in the debate surrounding Meta AI and ChatGPT.
This multi-platform approach allows ARM to diversify its revenue streams and technology destinations, with a licensing-focused model that has proven its scalability for years. Each new design adapted for AI, PCs, or mobile devices generates additional revenue without requiring direct intervention in production.
With an attractive combination of energy efficiency, design flexibility, competitive cost, and support from major technology partners, ARM's strategy appears well-aligned with the needs of modern computing. 2025 will be a key year to see to what extent this vision can be translated into concrete market results. The momentum ARM is gaining in a landscape increasingly focused on AI and sustainability cannot be underestimated. The growing participation of leading companies like AWS, Microsoft, and Google, coupled with the appeal of NVIDIA products and the versatility of its business model, positions ARM as an increasingly relevant player, capable of shaping the future architecture of data centers.
Passionate writer about the world of bytes and technology in general. I love sharing my knowledge through writing, and that's what I'll do on this blog, show you all the most interesting things about gadgets, software, hardware, tech trends, and more. My goal is to help you navigate the digital world in a simple and entertaining way.