Arm Says Data Center Will 'Soon' Be Its Top Revenue Driver

Arm Holdings, the chip architecture company that powers virtually every smartphone on Earth, says its data center business will 'soon' leap ahead to become its single largest revenue engine — a seismic shift for a company long defined by mobile and low-power computing.

The announcement marks a pivotal inflection point in the semiconductor industry, as hyperscale cloud providers like Amazon Web Services (AWS) and Microsoft Azure increasingly turn to Arm-based custom silicon to power their sprawling server fleets, challenging the decades-long dominance of Intel and AMD in the data center.

Key Takeaways

• Arm expects data center revenue to surpass smartphone revenue as its largest business segment
• Amazon (Graviton) and Microsoft (Cobalt) already deploy custom Arm-based data center processors
• The shift reflects Arm's successful pivot from low-power mobile chips to high-performance server computing
• Arm's architecture now competes directly with x86 designs from Intel and AMD in enterprise workloads
• AI infrastructure demand is accelerating Arm's data center adoption
• Arm's royalty-based business model means higher-value server chips generate significantly more revenue per unit

From Smartphones to Server Racks: Arm's Strategic Evolution

Arm's journey to data center prominence has been years in the making. The Cambridge, UK-based company originally built its reputation on energy-efficient processor architectures designed for battery-powered devices — think early PDAs, feature phones, and eventually the smartphones that now define modern life.

Today, Arm-based chips power more than 99% of the world's smartphones. Every iPhone runs on Apple's custom Arm-based silicon. Nearly every Android device does the same. But smartphone market saturation and fierce pricing pressure have pushed Arm to look beyond mobile for its next growth chapter.

The data center represents that opportunity. Unlike smartphones, where Arm earns relatively modest royalties on each chip sold, server-grade processors command significantly higher price points. A single data center CPU can cost thousands of dollars, compared to a mobile chip that might cost $20 to $50. That pricing differential means Arm's per-unit royalty revenue from data center chips dwarfs what it earns from mobile.

Amazon and Microsoft Lead the Arm Server Revolution

The biggest validation of Arm's data center ambitions comes from 2 of the world's largest cloud providers. AWS has been the most aggressive mover, developing multiple generations of its Graviton processor family. The latest Graviton4 chips, launched in 2024, deliver what Amazon claims is up to 30% better price-performance than previous generations and competitive performance against top-tier x86 alternatives.

Microsoft followed with its own custom Arm-based server chip, Cobalt 100, which began rolling out across Azure data centers in 2024. The 128-core processor targets general-purpose cloud workloads and represents Microsoft's first in-house server CPU.

Other major players are also investing heavily in Arm-based data center silicon:

• Google has explored Arm-based designs for internal workloads
• NVIDIA pairs its Grace CPU (Arm-based) with Hopper GPUs for AI supercomputing
• Ampere Computing sells Arm server chips to cloud providers and enterprises
• Oracle offers Arm-based instances on Oracle Cloud Infrastructure
• Alibaba and other Chinese cloud giants have developed Arm-based server processors

This broad industry adoption signals that Arm's data center play is not a niche experiment — it is becoming mainstream infrastructure.

Why AI Is Accelerating Arm's Data Center Growth

The explosive growth of artificial intelligence workloads has created an unexpected tailwind for Arm in the data center. While GPUs from NVIDIA dominate AI training, the surrounding infrastructure — including the CPUs that manage data pipelines, orchestrate workloads, and handle inference tasks — increasingly runs on Arm-based processors.

NVIDIA's Grace Hopper Superchip architecture pairs an Arm-based Grace CPU directly with an H100 or H200 GPU, creating a tightly integrated system optimized for AI workloads. This design choice by the world's most valuable semiconductor company sends a powerful signal about Arm's suitability for high-performance computing.

Energy efficiency also plays a critical role. Data centers consume enormous amounts of electricity, and AI workloads are making the problem worse. Arm's architectural DNA — born from decades of optimizing for power-constrained mobile devices — gives it a natural advantage. Arm-based server chips typically deliver better performance-per-watt than comparable x86 designs, which translates directly into lower electricity bills and reduced cooling requirements for data center operators.

With global data center power consumption projected to double by 2026 according to the International Energy Agency (IEA), energy efficiency is no longer a nice-to-have — it is a business imperative.

The x86 Incumbents Face Growing Pressure

Arm's rise in the data center comes at a challenging time for traditional x86 chip makers. Intel, which once held more than 90% of the server processor market, has seen its share erode steadily. The company's manufacturing struggles, delayed product roadmaps, and leadership changes have created openings that both AMD and Arm have exploited.

AMD has gained significant data center market share with its EPYC processor line, but now faces a 2-front competitive battle — defending against Intel's comeback attempts while also contending with Arm's architectural incursion.

The competitive dynamics break down along several key dimensions:

• Software ecosystem: x86 has decades of enterprise software optimization, but Arm compatibility has improved dramatically
• Performance: Arm designs now match or exceed x86 in many workloads, particularly cloud-native applications
• Power efficiency: Arm maintains a structural advantage in performance-per-watt
• Custom silicon: Arm's licensing model lets cloud providers design chips tailored to their specific workloads — something x86 vendors cannot easily replicate
• Cost: Hyperscalers building their own Arm chips can potentially undercut third-party x86 pricing

The custom silicon angle is particularly significant. When AWS builds a Graviton chip, it can optimize every aspect of the design for its own cloud workloads, eliminating features it does not need and enhancing those it does. This level of customization is simply not possible when buying off-the-shelf x86 processors from Intel or AMD.

Arm's Business Model Amplifies the Revenue Impact

Understanding why data center growth matters so much to Arm requires examining the company's unique business model. Unlike Intel or AMD, Arm does not manufacture or sell chips. Instead, it licenses its processor architecture to chip designers and collects royalties on every chip shipped.

Arm's revenue comes from 2 primary streams: upfront licensing fees paid by companies designing new chips, and per-unit royalties collected when those chips ship. The royalty rate is typically a small percentage of the chip's selling price.

This is where the math gets interesting. A smartphone application processor might sell for $30 to $50, generating a royalty of perhaps $1 to $2 for Arm. A data center processor might sell for $3,000 to $10,000, generating royalties that could be 50 to 100 times higher per unit.

Even if Arm ships far fewer data center chips than smartphone chips, the revenue contribution can be enormous. Arm has also been pushing to capture a larger royalty share through its newer Arm Total Access and Arm Flexible Access licensing programs, which bundle more intellectual property and command higher fees.

Since its IPO on the Nasdaq in September 2023 — which valued the company at roughly $54 billion — Arm's stock has surged, reflecting investor enthusiasm about precisely this data center growth story. The company's market capitalization has at times exceeded $150 billion, making it one of the most valuable semiconductor companies globally.

What This Means for the Industry

Arm's data center ascent carries broad implications across the technology landscape. For cloud providers, it validates the strategy of investing in custom silicon rather than relying solely on merchant chip vendors. For enterprise IT teams, it means evaluating Arm compatibility for workloads that may eventually run on Arm-based cloud instances.

For developers, the message is clear: Arm is no longer just a mobile platform. Applications targeting cloud deployment should be tested and optimized for Arm architectures. Major software ecosystems — including Linux distributions, container platforms like Docker and Kubernetes, and databases like PostgreSQL and MySQL — already run well on Arm.

For Intel and AMD, Arm's growth represents a structural competitive threat that licensing agreements and architectural tweaks alone cannot address. Both companies will need to accelerate innovation in power efficiency and offer more flexible engagement models to retain hyperscale customers.

Looking Ahead: Arm's Path to Data Center Dominance

Arm has not provided a specific timeline for when data center revenue will overtake smartphones, but the trajectory appears clear. Several factors suggest the crossover could happen within the next 2 to 3 years.

AWS continues to expand Graviton adoption across its instance portfolio. Microsoft is scaling Cobalt 100 deployment. New entrants continue to announce Arm-based server designs. And the AI infrastructure buildout shows no signs of slowing.

The key risk for Arm is RISC-V, an open-source processor architecture that some view as a long-term competitive threat. Unlike Arm, RISC-V does not require licensing fees, making it attractive to cost-conscious chip designers. However, RISC-V's data center ecosystem remains immature, and Arm's head start in software compatibility, performance validation, and customer relationships provides a substantial moat.

If Arm's data center business does become its top revenue driver, it will complete one of the most remarkable strategic pivots in semiconductor history — transforming from a mobile-first architecture company into the foundation of global cloud and AI infrastructure.