Arm Says Data Center Will Be Its Biggest Business by 2030

Arm Holdings just delivered a record-breaking quarter — and a bold declaration about the future of computing. The chip architecture giant reported fiscal Q4 2026 revenue of $1.49 billion, up 20% year-over-year, while CEO Rene Haas told investors that data center is on track to become the company's largest business segment, with Arm poised to capture 'the largest share of the CPU market' by 2030.

The results cap a fiscal year in which Arm generated $4.92 billion in total revenue, marking its third consecutive year of 20%-plus growth — a remarkable feat for a company whose architecture now underpins everything from smartphones to the world's most powerful AI servers.

Key Takeaways

• AGI CPU demand exploded: Customer orders doubled from $1B to $2B in just 6 weeks after the product's launch
• Data center royalties doubled year-over-year, with management expecting them to double again
• Q4 FY2026 revenue: $1.49B (+20% YoY), beating the midpoint of prior guidance
• Full-year FY2026 revenue: $4.92B (+23% YoY), a new annual record
• Long-term target reaffirmed: $25B in revenue by FY2031
• Supply constraints are the primary bottleneck, not demand — Arm is racing to expand wafer, packaging, and testing capacity

AGI CPU Demand Doubles to $2 Billion in Just 6 Weeks

The headline number from Arm's earnings call was staggering. The company's Arm AGI CPU, unveiled at the 'Arm Everywhere' event in late March, has seen customer demand surge from an initial $1 billion supply commitment to over $2 billion in committed orders — all within 6 weeks of launch.

This is not vaporware demand. Haas emphasized that these figures represent actual customer commitments, driven by the escalating compute requirements of AI agents and the infrastructure needed to support them. The AGI CPU is designed specifically for the intersection of artificial general intelligence workloads and data center efficiency, a market segment that barely existed 3 years ago.

However, Arm is keeping its annual revenue guidance for the AGI CPU product at $1 billion. The reason is not a lack of appetite from hyperscalers and cloud providers — it is a supply problem. The company is 'working at full capacity' to expand wafer production, advanced packaging, and testing infrastructure to meet the tidal wave of orders.

This supply-demand gap underscores a broader industry dynamic: the AI infrastructure buildout is moving faster than the semiconductor supply chain can keep up, even for companies that design chips rather than fabricate them.

Data Center Poised to Become Arm's Largest Business

For decades, Arm was synonymous with mobile. The company's architecture powers virtually every smartphone on the planet, generating steady royalty streams from billions of devices shipped annually. That era of mobile dominance is not ending — but it is being eclipsed.

Data center royalty revenue doubled year-over-year in Q4, and Haas made a striking prediction: it will double again. When that happens, data center will surpass mobile to become Arm's single largest revenue segment.

This transformation has been years in the making. Amazon Web Services launched its custom Arm-based Graviton processors in 2018, proving that Arm architectures could compete with x86 in server workloads. Since then, Microsoft Azure, Google Cloud, and Oracle Cloud have all introduced or announced Arm-based server chips. NVIDIA's Grace CPU, built on Arm architecture, is now a cornerstone of the company's AI data center platform.

The shift is being accelerated by AI. Large language models and inference workloads demand massive parallelism and energy efficiency — areas where Arm's architecture has inherent advantages over traditional x86 designs. As AI inference moves from centralized GPU clusters to distributed edge-to-cloud deployments, Arm's power-per-watt advantage becomes increasingly compelling.

The $25 Billion Revenue Target: Ambitious but Credible

Arm reaffirmed its long-term goal of reaching $25 billion in annual revenue by FY2031 — roughly 5x its current run rate. At first glance, this target seems extraordinarily aggressive. But the math starts to make sense when you break down the growth vectors.

• Rising royalty rates: Arm has been steadily increasing its per-chip royalty through premium architecture licenses like Armv9, which commands significantly higher fees than its predecessor
• Data center expansion: If data center royalties continue doubling annually, this segment alone could contribute $8-10B by FY2031
• Automotive growth: Arm-based chips are becoming standard in software-defined vehicles, with design wins at major automakers
• AI PC and edge AI: The shift toward on-device AI processing in laptops and IoT devices creates new royalty streams
• AGI CPU direct sales: The new product line adds a hardware revenue stream that did not exist before FY2026
• Custom silicon proliferation: More companies are designing custom Arm-based chips, each generating licensing and royalty fees

Compared to Intel's peak revenue of roughly $79 billion in 2021 or AMD's $23 billion in 2024, Arm's $25 billion target would position it as one of the semiconductor industry's largest players by revenue — remarkable for a company that historically earned pennies per chip through licensing.

Why This Matters for the Broader AI Industry

Arm's results are not just a corporate earnings story. They signal a fundamental shift in the AI infrastructure stack that affects every major technology company.

For hyperscalers, the data is clear: Arm-based server architectures are no longer experimental alternatives to x86. They are becoming the default for new AI workloads. Amazon, Microsoft, and Google are all investing heavily in custom Arm silicon, and Arm's surging data center royalties confirm that these chips are shipping at scale.

For NVIDIA, the relationship with Arm remains symbiotic but complex. NVIDIA's Grace CPU relies on Arm architecture, and the two companies collaborate closely on AI data center platforms. But Arm's growing ambition in the data center — particularly with its own AGI CPU — could introduce competitive tension over time.

For Intel and AMD, Arm's data center momentum represents an existential competitive threat. Intel has already lost significant server market share to AMD over the past 5 years. Now both x86 incumbents face a third front: Arm-based alternatives that offer better performance-per-watt for AI inference and cloud-native workloads.

The AI agent revolution adds another dimension. As AI agents become more sophisticated, they require orchestration across multiple compute tiers — cloud, edge, and device. Arm's presence across all 3 tiers gives it a unique architectural advantage that neither Intel nor AMD can easily replicate.

Supply Chain Bottlenecks Remain the Wild Card

Despite the demand surge, Arm's near-term growth is constrained by physical reality. The company acknowledged that wafer supply, advanced packaging, and testing capacity are all limiting factors for the AGI CPU product line.

This is a familiar challenge across the AI hardware ecosystem. TSMC, which manufactures the vast majority of advanced Arm-based chips, is already operating at near-full capacity for its cutting-edge process nodes. Advanced packaging technologies like CoWoS (Chip-on-Wafer-on-Substrate) remain in short supply, creating bottlenecks for complex multi-chiplet designs.

Arm is not a chip manufacturer — it licenses its architecture to companies that design and fabricate silicon. But the AGI CPU represents a strategic shift toward offering more complete chip solutions, which means Arm is now directly exposed to supply chain constraints that it previously avoided.

The company's decision to maintain a conservative $1 billion revenue guidance for the AGI CPU despite $2 billion in demand suggests management is being prudent about supply chain risks. If Arm can successfully scale production, the upside could be substantial.

Looking Ahead: The Race to 2030

Arm's vision for 2030 is sweeping: the largest share of the global CPU market, a dominant position in data center computing, and a revenue base 5 times larger than today. The company is betting that AI will reshape the entire compute landscape in its favor.

Several milestones to watch in the coming quarters:

• AGI CPU production ramp: Can Arm scale beyond the $1B supply constraint by mid-FY2027?
• Data center royalty trajectory: Will the doubling trend continue as more hyperscalers deploy Arm servers?
• Competitive response from x86: Intel's upcoming Clearwater Forest and AMD's next-gen Turin server chips will test whether x86 can defend its data center share
• Custom silicon adoption: Watch for new Arm-based chip announcements from major cloud providers and AI companies
• Automotive and edge AI growth: These segments could become significant revenue contributors by FY2028

The semiconductor industry has seen bold predictions before — and not all of them have materialized. But Arm's combination of surging demand, architectural advantages in AI workloads, and an expanding addressable market makes its 2030 ambitions more credible than most.

For investors, developers, and enterprise IT leaders alike, the message from Arm's earnings call was unmistakable: the era of x86 dominance in the data center is ending, and Arm intends to be the primary beneficiary of what comes next.