Arm CEO Says Data Center Business Is 'Exploding'
TL;DR
Arm reported record quarterly revenue of $1.24 billion driven by 100% year-over-year growth in data center royalties, with the segment poised to surpass smartphones as the company's largest business within a few years amid massive AI infrastructure investments.
📊 Financial Performance & Diversification 2 insights
Record revenue and royalties beat expectations
Arm generated $1.24 billion in revenue and nearly $740 million in royalties, representing 27% year-over-year growth despite broader semiconductor headwinds.
Memory shortage impact minimized by business mix
The company avoided significant impact from smartphone memory constraints because lower-end devices affected by shortages carry lower royalty rates, while high-growth data center revenues offset any weakness.
🖥️ Data Center Architecture Revolution 3 insights
Data center royalties doubled year-over-year
Data center revenue grew over 100% annually and is expected to become Arm's largest business segment sooner than anticipated, potentially within a few years.
Core count expansion drives royalty acceleration
Hyperscaler chips are adopting more ARM cores per processor, jumping from 96 to 192 cores in AWS Graviton and 72 to 88 in NVIDIA Vera, directly increasing royalty payments per chip.
Agentic AI workloads require CPU capabilities
As AI evolves toward agentic workflows that manage complex operations across systems, these tasks require CPU processing that GPUs cannot handle, cementing ARM's role in heterogeneous compute environments.
🤝 AI Infrastructure & Strategic Partnerships 3 insights
Massive CapEx validates AI build-out sustainability
Hyperscalers including Alphabet, Meta, and Microsoft are investing billions beyond initial projections—rendering the $500 billion Stargate initiative credible—and confirming AI infrastructure spending is not a bubble but a long-term transformation.
NVIDIA partnership deepens with standalone Vera CPU
NVIDIA's plan to sell Vera CPUs standalone, using six times more ARM cores than previous Grace generations for storage and offload functions, accelerates adoption of homogeneous ARM-based data center architectures.
SoftBank reaffirms long-term commitment
CEO Masayoshi Son has no intention of selling SoftBank's stake in Arm, viewing the company as significantly undervalued relative to its growth trajectory in AI and data center markets.
Bottom Line
Arm's programmable, low-power architecture positions it to capture increasing value per chip as data centers transition to ARM-native, high-core-count CPU configurations to handle agentic AI workloads, making the data center segment the company's primary growth engine within three years.
More from Bloomberg Technology
View all
CoreWeave Shares Drop After Forecast Sparks Growth Fears | Bloomberg Tech 5/8/2026
CoreWeave defended its growth trajectory amid investor skepticism over compressed margins, while AI automation drives the worst job market for Gen Z in 37 years and retail investors fuel a 6,000% surge in AST SpaceMobile shares.
Arm Warns of Phone Market Weakness | Bloomberg Tech 5/7/2026
Arm Holdings reports that booming AI data center demand is offsetting smartphone market weakness, while the tech sector faces a three-year high in layoffs as companies cut jobs to fund AI infrastructure investments. Strategic alliances like Anthropic accessing SpaceX compute highlight the scramble for processing capacity, and IonQ advances toward fault-tolerant quantum computing.
AMD Soars on Blockbuster AI-Fueled Forecast | Bloomberg Tech 5/6/2026
AMD shares surged 15% on a blockbuster AI forecast with data center CPU growth hitting 70%, while Nvidia secured optical fiber supply chains via a $500M strategic investment; meanwhile, Uber and Disney beat earnings expectations through diversification, though Microsoft's climate goals face pressure from surging AI energy demands.
Apple Weighs Using Intel, Samsung Processors | Bloomberg Tech 5/5/2026
Apple is exploring U.S.-based chip manufacturing with Intel and Samsung to reduce reliance on TSMC amid geopolitical tensions, while major AI firms expand government oversight agreements and Pinterest demonstrates success with specialized, cost-efficient AI models.