Arm Holdings delivered a first-quarter revenue and earnings forecast that cleared Wall Street expectations by a wide margin, fueled by surging demand for its energy-efficient chip architectures inside AI data centers. But a persistent slump in the smartphone market weighed on royalty income, and shares dropped as much as 9% in extended trading as investors struggled to reconcile the two storylines.
The British chip designer projected adjusted earnings per share of 40 cents for the fiscal first quarter, roughly 11% above the 36-cent consensus tracked by Reuters. Revenue guidance of approximately $1.24 billion also came in above the analyst consensus of roughly $1.18 billion, according to the same Reuters report, though the company did not break out how much of the upside stemmed specifically from data center licensing versus other segments.
“We are seeing unprecedented demand for Arm’s compute platform across the AI ecosystem,” CEO Rene Haas said on the company’s earnings call on May 6, 2026. Haas pointed to new licensing agreements with hyperscale cloud operators as evidence that Arm-based server processors are moving from experimental deployments into production-scale adoption. “The energy efficiency of our architecture is no longer just a mobile advantage. It is becoming a requirement for anyone building AI infrastructure at scale,” he added.
Hyperscale cloud operators, facing electricity bills that can rival the cost of the hardware itself, have increasingly turned to Arm-based server processors that deliver more computation per watt than traditional x86 alternatives. Amazon’s Graviton chips, Nvidia’s Grace CPU, and Microsoft’s Cobalt processor all run on Arm architectures, giving the company a foothold inside three of the world’s largest cloud platforms.
The AI data center tailwind
Arm does not fabricate silicon. It licenses instruction-set architectures and core designs to chipmakers, then collects royalties on every unit shipped. That model means even small gains in data center market share can compound quickly: each new server processor that ships with an Arm core generates a per-unit royalty payment, and AI training clusters can contain thousands of processors per facility.
The economics are straightforward. Power and cooling represent a significant portion of data center operating costs, and those costs are climbing as AI workloads grow more demanding. Arm’s architectures, originally engineered for battery-constrained smartphones, are built around performance-per-watt efficiency. That engineering DNA now gives Arm a structural advantage in a market where every watt saved translates directly into lower operating expenses for cloud providers.
The EPS beat suggests this advantage is moving beyond pilot deployments and into production-scale adoption. Licensing revenue, which reflects new design wins, and royalty revenue, which reflects chips already shipping, both feed into the forecast. A beat of this size indicates that at least one of those streams, and likely both, is tracking ahead of what analysts had modeled.
Smartphone weakness clouds the picture
The other half of the story is less encouraging. Arm’s royalty revenue took a hit from softness in the global smartphone market, according to Bloomberg, which reported that the company’s sales forecast failed to satisfy investors looking for a larger AI payoff. According to industry tracker Counterpoint Research, Arm-based processors power more than 99% of the world’s smartphones, which means the company’s royalty line is tightly coupled to handset shipment volumes.
The weakness is not new. Global smartphone shipments have been uneven for several quarters, pressured by cautious consumer spending in key markets and inventory adjustments among handset manufacturers. What makes it notable now is the contrast with the AI data center momentum. Investors had hoped the data center upside would be large enough to fully offset the mobile drag, and the mixed result left some disappointed.
The nature of the smartphone slowdown matters for how quickly royalties could recover. A temporary inventory correction among manufacturers like Samsung or Xiaomi might unwind within a quarter or two. A longer-term shift toward lower-priced devices, which carry smaller royalty rates per chip, could compress Arm’s mobile yield for a more extended period. The company’s guidance did not provide enough detail to distinguish between those scenarios.
Why shares fell despite the beat
Arm’s stock had rallied sharply over the past year on expectations that AI would transform the company’s growth profile. That rally pushed the valuation to levels that left little room for anything short of a blowout quarter. A forecast that beat consensus but came paired with smartphone headwinds was not enough to sustain the momentum.
The 9% after-hours decline, as reported by Reuters, reflects that tension. Some of the selling likely represents profit-taking by investors who rode the AI narrative higher and saw the mixed results as a reason to trim positions. After-hours trading is also thinner and more volatile than regular sessions, which can amplify moves in either direction.
Stacy Rasgon, a semiconductor analyst at Bernstein, noted that Arm’s valuation had already priced in aggressive AI-driven growth. “The beat was real, but the bar was extraordinarily high going into the print,” Rasgon told Reuters. “Investors wanted to see the data center story overwhelm everything else, and it did not quite get there this quarter.”
For longer-term shareholders, the more relevant question is whether Arm can keep converting its architectural advantages into durable revenue across multiple end markets. The company’s design wins now span data center processors, automotive chips, and Internet of Things devices, in addition to its dominant mobile franchise. If the data center ramp continues at the pace the latest guidance implies, the smartphone drag becomes a smaller share of the overall business with each passing quarter.
What the guidance does and does not reveal
The 40-cent EPS forecast and the roughly $1.24 billion revenue outlook are the hardest data points from this earnings cycle, and they directly support the case that Arm’s business is outrunning analyst models. But several gaps remain. The company has not disclosed a detailed revenue breakdown by end market, making it difficult to measure exactly how fast data center licensing is growing relative to mobile royalties. Independent benchmarks comparing Arm-based server chips against x86 alternatives from Intel and AMD in current AI workloads were not cited in the earnings materials, leaving the efficiency argument supported by industry consensus rather than fresh, quarter-specific data.
There is also limited visibility into the lag between new licensing deals and the royalty revenue they eventually produce. Data center chip designs can take years to move from initial license to volume production, which means some of the deals driving today’s optimism may not show up in royalty figures until well after the current fiscal year. Management’s confidence in the trajectory is clear from the guidance, but the precise timing remains opaque.
Arm’s data center bet measured against its mobile roots
What this earnings cycle makes plain is that Arm is in the middle of a structural shift. The company that built its business powering smartphones is becoming a foundational supplier to AI infrastructure, and the financial results are starting to reflect that change. The transition is uneven. Quarters where smartphone softness collides with strong data center demand will produce results that look messy even when the underlying trajectory is positive.
For now, the AI data center opportunity is large enough to pull Arm’s earnings above expectations, but not yet large enough to make the smartphone business irrelevant. How quickly that balance tips will determine whether the current stock price looks like a bargain or a peak when investors look back on this period a year from now.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
